40m QIL Cryo_Lab CTN SUS_Lab TCS_Lab OMC_Lab CRIME_Lab FEA ENG_Labs OptContFac Mariner WBEEShop
  PSL, Page 3 of 52  Not logged in ELOG logo
ID Date Author Typedown Category Subject
  1035   Tue Aug 7 23:30:19 2012 ranaSummaryEnvironmentEQ

Just felt an EQ. Impulse moved some vertical blinds by several mm.

Tue Aug 07 23:26:06 2012 

  1088   Mon Jan 7 13:54:29 2013 KristenSummaryPMCModal Frequency Testing for PMC

The PMC was tested & lowest resonant frequency was 330 Hz;  FEA model was adjusted to new frequency of 441 Hz

Results from December 18, 2012

The PMC in 058B W. Bridge was secured with several dog clamps to the laser table.  This table is not as stiff as the table in the Modal Lab in Downs, but was thought to be sufficient for this test.  Testing was done with the B&K system, using a laser vibrometer for the accelerometer and the small 8206 B&K hammer for excitation.  Below is a representation of the axis for this test, to understand where the PMC was excited and measured.

PMC_Measurement_Diagram.JPG

 

Measurements and excitations were approximately at the center of the corresponding face, as indicated in the image.

Below is a graph of the results of these measurements.  You can see that the lowest resonant frequency is at 330 Hz.

Experimental_Modal_Data_-_Dec_18.JPG 

 

Next Steps:

Next I will update the ANSYS model to be more accurate, hopefully showing about 330 Hz as the lowest mode.

FEA Model in ANSYS:

Briefly: the previous model reported in the elog was changed by refining the mesh on the slots/cones and the bearings.  This would allow for that portion to behave more accurately.  The contacts were left as Program-controlled (any other control seemed to overestimate the contact, raising the predicted resonant frequencies). 

Below is an image of the lowest mode, at 441 Hz.  The arrow indicates the motion - the mode is roughly a transverse flagpole mode.

ANSYS_PMC_Mode_441.JPG

Now that the model has been made more accurate, steps can be taken to raise the resonant frequency. While the initial goal that was mentioned to me of 1kHz is improbable, there are certainly ways to raise the frequency and damp the modes that are problematic.

 

 

  1089   Wed Jan 16 16:48:39 2013 ranaSummaryPMCModal Frequency Testing for PMC

  We are interested in the longitudinal mode along the Y direction. That is the only one which is problematic for the servo. Please remeasure so that you excite Y and measure Y and then model the first longitudinal mode.

The other modes are interesting, but they're not the main thing we care about.

  1205   Sun Jun 23 11:07:27 2013 ranaSummaryTempCtrlTemperature Noise plots: sensors

 Frank gave us this data about temperature sensors some time ago. The first plot compares the low frequency performance of a few sensors. The AD590 is the worst performer, although perhaps the easiest to use. The second plot shows several kinds of temperature noise compared with a 100 K NTC sensor.

tempsensnoise2.pdftempnoise_final.pdf

The AD590 is an OK sensor for the outside of the can since out loop gain is not likely to get to this sensor noise performance. To monitor the temperature the noise inside the vacuum, however, we would have to use one of the better ones. 

http://en.wikipedia.org/wiki/Resistance_thermometer

Attachment 2: tempnoise_final.pdf
tempnoise_final.pdf
  1336   Thu Sep 12 18:12:26 2013 Matt A.SummaryNoiseBudgetCoating Thermal Noise Calculator
As part of checking Tara's code for optimizing the coating structure to reduce thermo-optic noise, I wrote this coating thermal noise calculator with all the values needed for calculating the Thermo-elastic, Thermo-refractive, Thermo-optic, and Brownian noise PSD of an GaAs/AlGaAs coating. 

Tara's code is a bastardized version of GWINC, and as such, it's horrible to read, and who knows where there might be errors. 

This code is based on the same papers as GWINC, but written semi- independently. That is: I wrote the code directly from the papers that are referenced in GWINC, and the code was checked against the output of GWINC, but none of the code in GWINC is directly copied, and in some cases, my code is based on the published papers instead of the ArXiv versions which are sometimes less clear. 

While this is currently just one long function, it is heavily commented and referenced, and can be easily augmented into a general-purpose coating thermal-noise calculator.

The code is called CoatingThermalNoiseCalc.m, and it calls only one additional function: the besselzero.m function that GWINC uses to calculate the zeros of a Bessel function.
It takes three inputs: x, firstN, and f. 

'x' is an array of values with each index representing the optical thickness of one layer in units of the laser wavelength, with the first index representing the topmost layer, and the last index representing the layer before the that is against the substrate. Therefore, if you wanted to calculate a 55-layer quarter-wave stack, you could use x = 1/4*ones(1,55); 
If you wanted to use a 1/2 layer cap on top of that, you could use x = [1/2,1/4*ones(1,55)]; etc.

'firstN' in an indicator of whether the top layer has a high index of refraction, or a low index. Therefore, it takes only two inputs: 'nH' or 'nL' for high index or low index, respectively. This entry is not case sensitive.

'f' is an array of frequencies in HZ for which you'd like to calculate the noise. This is generally generated using the logspace command: f = logspace(0,4,1000); would generate an array of 1000 frequencies logarithmically spaced from 1 to 10^4 Hz.

The code returns a structure containing the input frequency, .f, the calculated reflectivity of the layer stack, .Refl, the Transmittivity of the stack, .Trans, (which should equal 1-Refl) the phase of the reflected beam at the surface, .Rphase, the incoherent thermo-elastic noise psd, .SteZ, the incoherent thermo-refractive noise psd, .StrZ, the coherently added thermo-elastic and thermo-refractive noise, called the thermo-optic noise, .StoZ, and the Brownian thermal noise psd, .SbrZ. Where the power spectral densities are in units of m^2/Hz. These noises are for a single mirror with the mirror structure given by x, and currently, the laser wavelentgh is 1064 nm, but that is easily changed in the code, or the code can be easily re-written to accept further inputs. 

All the necessary code is stored in https://nodus.ligo.caltech.edu:30889/svn/trunk/mattabe/CoatingThermalNoiseCalc/
Along with an example.

I've checked the output against that of GWINC in three different cases:

Case 1: two 1/4-wavelength layers
Case 2: 55 1/4-wavelength layers
Case 3: Tara's optimized 55-layer coating

The attached figures show that the two techniques agree to better than 20% of the GWINC output, with most of the mismatch at higher frequencies. I'm not yet sure why that is. It could be due to my improved integration in calculating Evans' thick coating correction, it could be due to my using a different form of the equation for Braginsky's finite substrate correction for the thermo-elastic noise, or it could just be due to some minor differences in the precision of some of the input values. 

If you would like to run the code yourself, and you have any questions, let me know. I've also got the values you would need for calculating silica/tantala coatings.
 

 

Attachment 1: Case1_Noise.pdf
Case1_Noise.pdf
Attachment 2: Case1_Resid.pdf
Case1_Resid.pdf
Attachment 3: Case2_Noise.pdf
Case2_Noise.pdf
Attachment 4: Case2_Resid.pdf
Case2_Resid.pdf
Attachment 5: Case3_Noise.pdf
Case3_Noise.pdf
Attachment 6: Case3_Resid.pdf
Case3_Resid.pdf
  1337   Fri Sep 13 13:40:43 2013 ranaSummaryNoiseBudgetCoating Thermal Noise Calculator

 combined PDFs with pdftk:

pdftk *.pdf cat output MattThermal.pdf

and saved as PDF-X for thumbnail compatibility

Attachment 1: MattThermal.pdf
MattThermal.pdf MattThermal.pdf MattThermal.pdf MattThermal.pdf MattThermal.pdf MattThermal.pdf
  1338   Tue Sep 17 19:43:45 2013 taraSummaryNoiseBudgetCoating Thermal Noise Calculator

Quote:
...
...
The attached figures show that the two techniques agree to better than 20% of the GWINC output, with most of the mismatch at higher frequencies. I'm not yet sure why that is. It could be due to my improved integration in calculating Evans' thick coating correction, it could be due to my using a different form of the equation for Braginsky's finite substrate correction for the thermo-elastic noise, or it could just be due to some minor differences in the precision of some of the input values. ..

 

 

 I checked the calculation. I think most of the discrepancies are from the thick coating correction calculation (from Evans etal paper). The error is frequency dependent, and the calculations that involve frequency dependence are temperature fluctuation and thick coating correction. The temperature fluctuations are the same from our results. So it is most likely the thick coating correction. I checked and the corrections did differ at high frequency.

 I need to take a closer look to tell exactly where the errors are. Since the error is small and only at high frequency (around the shot noise limit, 10kHz),  I don't think it will be a problem for us.

  1342   Thu Sep 19 14:55:11 2013 taraSummaryNoiseBudgetCoating Thermal Noise Calculator

Quote:

Quote:
...
...
The attached figures show that the two techniques agree to better than 20% of the GWINC output, with most of the mismatch at higher frequencies. I'm not yet sure why that is. It could be due to my improved integration in calculating Evans' thick coating correction, it could be due to my using a different form of the equation for Braginsky's finite substrate correction for the thermo-elastic noise, or it could just be due to some minor differences in the precision of some of the input values. ..

 

 

 I checked the calculation. I think most of the discrepancies are from the thick coating correction calculation (from Evans etal paper). The error is frequency dependent, and the calculations that involve frequency dependence are temperature fluctuation and thick coating correction. The temperature fluctuations are the same from our results. So it is most likely the thick coating correction. I checked and the corrections did differ at high frequency.

 I need to take a closer look to tell exactly where the errors are. Since the error is small and only at high frequency (around the shot noise limit, 10kHz),  I don't think it will be a problem for us.

Tara noticed an accidental re-definition in my old code. I fixed it, and updated the svn. This fixes most of the discrepancies, but shifts the difference in thermo-optic to the low-frequency region.

Attachment 1 is the comparison from case 3 between mine and Tara's calculations of his optimized coating structure.

Attachment 2 is the comparison from case 2 between mine and Tara's calculations of a 55-layer 1/4-wavelength stack.

Attachment 1: ThermalUpdateEmbed.pdf
ThermalUpdateEmbed.pdf ThermalUpdateEmbed.pdf
Attachment 2: ThermalUpdateEmbed2.pdf
ThermalUpdateEmbed2.pdf ThermalUpdateEmbed2.pdf
  1343   Thu Sep 19 18:09:18 2013 taraSummaryNoiseBudgetCoating Thermal Noise Calculator

Quote:

 

Tara noticed an accidental re-definition in my old code. I fixed it, and updated the svn. This fixes most of the discrepancies, but shifts the difference in thermo-optic to the low-frequency region.

Attachment 1 is the comparison from case 3 between mine and Tara's calculations of his optimized coating structure.

Attachment 2 is the comparison from case 2 between mine and Tara's calculations of a 55-layer 1/4-wavelength stack.

 I discussed the calculation with Matt. The error in TO noise is large because it is a fraction of something small. Mostly it comes from TE part. The error in TO noise appears large (10%-20%) because the TO level is small.  Otherwise, the rests are in good agreement, and I think we should be able to order soon. 

 Below, summary of the calculation, dTE is alpha_effective * coating thickness, dTO is beta effective * lambda. 0.2% difference in dTE and 0% difference in dTR can cause error upto 40% in dTO when dTE and dTR cancel each other really well. But this will be insignificant, since the final TO levels are still in the same magnitude.

  Matt Gwinc
dTE 8.161e-11 8.141e-11
dTR -8.11e-11 -8.11e-11
dTO (dTE+dTR) 4.87e-13 2.88e-13

 

The summary of the TO cancellation is in wiki page AlGaAs

  1393   Tue Dec 31 19:33:47 2013 taraSummaryNoiseBudgetbeat measurement

I got a chance to measure beat measurement. The noise budget is updated and contains all dominant noise traces.

 

== Beat measurement ==

beat_2013_12_24.png

1) at DC to 10Hz, the contribution is mostly from RIN driven Photothermal noise and a bit of seismic noise, a small peaks around 10Hz is probably from the stack, not the cavity sagging. The hump from DC to ~ 50Hz disappear when it is quiet. I think it is mostly scattered light associated with the seismic noise, not displacement noise due to the vibration.

2) 10Hz to 1kHz is pretty much Coating Brownian noise.

3) At 1kHz and above, it is PLL readout noise and residual frequency noise from the laser, where the gain cannot suppress enough noise. This is mostly from ACAV. The residual frequency noise = free running noise / (1+ OLGTF). The measurement of the open loop gain is explained below.

 

==TTFSS Loop characterization==

The OLG TF of TTFSS is measured up to 10MHz and compared with the calculation. The schematic explaning how TTFSS actuates on the laser is shown below.

TTFSS.png

The freqeuncy discriminator can be measured from the slope of the error signal (from Common out1) while scanning the laser. For RCAV Dv = 1/ (194 kHz/V) and 1/(164kHz/V) for ACAV. with 1mW input power.

The adjustable gain stage can be tuned by turning the dial knob. At 400, gain=1, and the gain changes by 10dB with every 250click.

The PZT actuator has a gain of  4.5MHz/V (measured), and the EOM actuator is 15mRad/V (or 15mHz/f  Hz/V) (taken from the spec sheet).

OLG measurement is taken:  RCAV OLG is measured and plotted against the theoretical approximation, see the below figure.

RCAV_OLG.png

above: RCAV OLG TF. Note: The calculation and the measurement do not include the integrator with corner frequency at 4.6kHz.

 

There are some problems with ACAV loop and I could not increase the gain up as much as it used to be and the UGF is around only 200kHz , but the measurement matches the calculation. Right now RCAV servo has a better loop performance.

ACAV_OLGTF.png

 The calculated OLG TF trace(green) should go down at 1MHz or above because of the opamps' bandwidth. I used ideal Op Amps in the simulation because I don't have some op amps in my liso library. I'll see if I can fix it.

  1395   Wed Jan 15 10:56:08 2014 EvanSummaryNoiseBudgetBeat measurement: residual

I've taken the total noise trace, interpolated it so that it uses the same frequency array as the measurement trace, and performed the quadrature subtraction of the two to get the residual. I've also converted the beat to single-cavity length noise by multiplying by Lλ/sqrt(2)c, with L = 3.7 cm.

Attachment 1: short_nb_residual.pdf
short_nb_residual.pdf
Attachment 2: short_nb_residual.fig
  1396   Fri Jan 17 12:46:29 2014 EvanSummaryNoiseBudgetNoise budget fitting: code fork

I've forked the noise budget code so that we can create a version that performs a fit to the coating loss angle. It is at CTNLab/simulations/noise_budget/iscmodeling/coating/iRefCav/nb_short_fit/nb_short_fit.m.

I've retooled the noise budget plot a bit. I've referred it to single-cavity length noise by multiplying the beat ASD by Lλ/(sqrt(2)c), where L = 3.7 cm. I've also combined some of the substrate noise, spacer noise, and technical noise traces so that there are not quite so many lines on a single plot. If we really want to display each trace individually, I think we should do so with a few separate plots (e.g., a thermal noise plot, a frequency/PLL noise plot, etc.). Fewer traces makes it easier for readers to make sense of the plot.

I'm going to start on writing the fitting code. For nonlinear least squares I'm used to using the Levenberg–Marquardt algorithm through scipy.optimize.curve_fit. I'll need to read up a bit on what's available in Matlab.

 

Attachment 1: nb_beat_short.pdf
nb_beat_short.pdf
  1397   Mon Jan 20 19:20:41 2014 EvanSummaryNoiseBudgetNoise budget fitting

I've added a χ2 minimization routine to nb_short_fit.m which looks for the value of ϕc (as defined by the Nakagawa formula) which makes the noise budget best fit the observed beat spectrum. For the weights in the χ 2 function, we need an estimate of the variance of the power in each bin. Ideally, we'd take multiple spectrum measurements and average them together. Since we only have a single measurement, for each bin I've taken the five bins on either side and computed the variance.

I performed the fit in the band from 26 Hz to 405 Hz because it looks like the total noise is dominated by coating Brownian noise in this region.

The first attachment shows χ2 as a function of ϕc. The routine assumes χ2 is parabolic in the neighborhood of the optimum value (which you can clearly see it is), and from this extracts the optimum value as well as the statistical uncertainty (which is given by the curvature of the parabola). From this the routine gives ϕc as 4.18(3) × 10−4, with a reduced of χ2 of 1.23.

From here, the next steps are

  • Settle on values for the lower and upper frequency limits for the fit. If the upper frequency is increased from 405 to 705 Hz, for example, the routine gives ϕc as 4.26(3) × 10−4, with a reduced of χ2 of 1.21. I think this is due (in part) to the fact that the noise budget is less than the measured noise near and above 1 kHz.
  • Perform the fit using the Harry formula, which includes the contributions from both loss angles rather than a single averaged value.
  • Compute the uncertainty arising from uncertainties in the other parameters. Tara has collected uncertainties for the material parameters in PSL:895, but we also need uncertainties in the spot size and the temperature.
Attachment 1: cost_func.pdf
cost_func.pdf
Attachment 2: nb_beat_short.pdf
nb_beat_short.pdf
Attachment 3: nb_beat_resid.pdf
nb_beat_resid.pdf
  1398   Wed Jan 22 16:18:37 2014 EvanSummaryNoiseBudgetNoise budget fitting: need uncertainties

In order to get the systematic uncertainty on ϕc, we need uncertainties in other parameters that enter the noise budget. Specifically:

  • Spot size, w. Currently using (182.0 ± 0.4) µm. The nominal comes from eq. 47 in Kogelnik and Li. This uncertainty can be propagated forward uncertainties in the following:
    • Cavity length, L. The spacer drawing (CTNLab/drawings/mechanical_drawings/dual_refcav/cavity_spacer_1.45inx1.5in.PDF) gives the length as (1.45 ± 0.01) inches.
    • Mirror radius, R. Nominal is 0.5 m; no idea about the uncertainty. It is not given in the test document. Currently using (500 ± 3) mm; i.e., an 0.5% uncertainty. This is the uncertainty claimed in the CVI catalog.
  • Coating thickness, d. Nominal is 4.5 µm; no idea about the uncertainty. Currently using (4.53 ± 0.07) µm from Tara's calculation (given in the reply to this elog post).
  • Substrate elastic modulus, Es. Using (73.1 ± 0.1) GPa as estimated from figure 29 in McSkimin 1953 (doi 10.1063/1.1721449), which is a (heavily cited) reference I found on the NIST ceramics database.
  • Substrate Poisson ratio, σs. Using 0.170 ± 0.004 as estimated from figures 29 and 30 in McSkimin (for Young and shear moduli, respectively) and then propagating the error forward to the Poisson ratio.
  • Cavity temperature. Should use 306(1) K.
  • Measurement uncertainty. Given an estimated PSD S(f) that is obtained by averaging M FFTs, the uncertainty should be S(f)/M1/2. Tara says to take M = 50. Need number of averages for each bin. Possibly need to adjust formula to account for Welch overlap.
  1400   Wed Jan 29 05:44:51 2014 taraSummaryNoiseBudgetNoise budget fitting: need uncertainties

I looked into the uncertainty in coating thickness of the QWL SiO2/Ta2O5 coating The thickness of  4.53 +/- 0.07 um (~1.5%)seems to be appropriate.

The thermal noise level is directly proportional to the coating thickness, so we want to estimate its uncertainty. The error in the thickness is from

  • The uncertainties in nL and nH: since the physical thickness is lambda/(4*n), the error in n goes to the error in d.
  • Manufacturing process.

The errors in nL and nH are quite small, nL ~ 1.45 +/ 0.01, nH ~ 2.06+/- 0.01. (From the literature). I also looked around the error in IBS thickness control, they are usually better than 0.1 nm, IBS, but that is the current technology. In literature around 2000s, 2% error seems to be the number estimated for the thickness control (Sullivan 2000, Badoil 2007). As a quick check, I used the same assumption for error propagation similar to that of AlGaAs coating. The result gives ~ 4.53 +/- 0.07 um for coating thickness.

Note that the error here is smaller than the difference in coating thickness for the coatings with or without half wave cap.

For 28 Layer (with cap), the coating thickness is 4.53 um,  for 28 layer QWL, the coating thickness is 4.35 um. But after digging up all the information from REO, and peter king they agree that it is 28 QWL with half wave cap.  I tried to compare the calculation and the photothermal TF measurement, but the effect is too small to be conclusive about the structure. So the biggest error might come from the fact that the coating has cap or not. The error is about 4%.

  1401   Wed Jan 29 21:36:53 2014 taraSummaryNoiseBudgeterror in spacer brownian noise

Spacer in BR noise

 

== COMSOL vs result from Kessler etal 2012==

     The analytical result from kessler2012, assume the force acts on whole surface of the spacer (with bore hole), I check this with COMSOL by comparing the result, similar to what I did in PSL:1075. The result agrees well within 2%. This verifies that COMSOL model is correct

 

==thermal noise level vs annulus thickness==


Typically, the contact surface between the spacer and the mirror is only a thin annulus, see psl:1199  . And the noise level is dependent on the actual area of contact. So I run the simulation to see the dependent of the stored energy (U) vs the annulus thickness. The annulus thickness is about 2 mm +/- 0.2 mm. The displacement noise is proportional to sqrt(U).

spacerBR_comsol.png

fig1: The stored energy as calculated by COMSOL, fitted with cubic polynomial.

The error from the contact area, the simulation result are small ~3% and 2%. These are smaller than the uncertainty of loss in bulk fused silica (can be from 10^-6 to 10^-7). The effect is still small in the total noise.

Attachment 2: spacerBR_comsol.fig
  1402   Fri Jan 31 00:44:47 2014 EvanSummaryNoiseBudgetNoise budget fitting: result

Quote:

In order to get the systematic uncertainty on ϕc, we need uncertainties in other parameters that enter the noise budget. Specifically:

  • Spot size, w. Currently using (182.0 ± 0.4) µm. The nominal comes from eq. 47 in Kogelnik and Li. This uncertainty can be propagated forward uncertainties in the following:
    • Cavity length, L. The spacer drawing (CTNLab/drawings/mechanical_drawings/dual_refcav/cavity_spacer_1.45inx1.5in.PDF) gives the length as (1.45 ± 0.01) inches.
    • Mirror radius, R. Nominal is 0.5 m; no idea about the uncertainty. It is not given in the test document. Currently using (500 ± 3) mm; i.e., an 0.5% uncertainty. This is the uncertainty claimed in the CVI catalog.
  • Coating thickness, d. Nominal is 4.5 µm; no idea about the uncertainty. Currently using (4.53 ± 0.07) µm from Tara's calculation (given in the reply to this elog post).
  • Substrate elastic modulus, Es. Using (73.1 ± 0.1) GPa as estimated from figure 29 in McSkimin 1953 (doi 10.1063/1.1721449), which is a (heavily cited) reference I found on the NIST ceramics database.
  • Substrate Poisson ratio, σs. Using 0.170 ± 0.004 as estimated from figures 29 and 30 in McSkimin (for Young and shear moduli, respectively) and then propagating the error forward to the Poisson ratio.
  • Cavity temperature. Should use 306(1) K.
  • Measurement uncertainty. Given an estimated PSD S(f) that is obtained by averaging M FFTs, the uncertainty should be S(f)/M1/2. Tara says to take M = 50. Possibly need to adjust formula to account for Welch overlap.

The above uncertainties are enough to estimate the statistical and systematic uncertainties on a fit to ϕc using the Nakagawa/Harry formula for a thin, lossy coating. By minimizing an appropriately weighted chi-squared function from 50 Hz to 500 Hz and then taking into account the above substrate and coating uncertainties, I find ϕc = (4.15 ± 0.03 stat ± 0.08 sys) × 10−4. More details will follow, and there may need to be some refinement (e.g., I still haven't dealt with the Welch overlap issue).

This has required adjusting the values of the Young modulus and Poisson ratio from their previous values (72.7 GPa and 0.167, respectively). I haven't checked these changes into the SVN.

  1404   Tue Feb 4 00:31:32 2014 EvanSummaryNoiseBudgetNoise budget fitting: result

Quote:

Quote:

In order to get the systematic uncertainty on ϕc, we need uncertainties in other parameters that enter the noise budget. Specifically:

  • Spot size, w. Currently using (182.0 ± 0.4) µm. The nominal comes from eq. 47 in Kogelnik and Li. This uncertainty can be propagated forward uncertainties in the following:
    • Cavity length, L. The spacer drawing (CTNLab/drawings/mechanical_drawings/dual_refcav/cavity_spacer_1.45inx1.5in.PDF) gives the length as (1.45 ± 0.01) inches.
    • Mirror radius, R. Nominal is 0.5 m; no idea about the uncertainty. It is not given in the test document. Currently using (500 ± 3) mm; i.e., an 0.5% uncertainty. This is the uncertainty claimed in the CVI catalog.
  • Coating thickness, d. Nominal is 4.5 µm; no idea about the uncertainty. Currently using (4.53 ± 0.07) µm from Tara's calculation (given in the reply to this elog post).
  • Substrate elastic modulus, Es. Using (73.1 ± 0.1) GPa as estimated from figure 29 in McSkimin 1953 (doi 10.1063/1.1721449), which is a (heavily cited) reference I found on the NIST ceramics database.
  • Substrate Poisson ratio, σs. Using 0.170 ± 0.004 as estimated from figures 29 and 30 in McSkimin (for Young and shear moduli, respectively) and then propagating the error forward to the Poisson ratio.
  • Cavity temperature. Should use 306(1) K.
  • Measurement uncertainty. Given an estimated PSD S(f) that is obtained by averaging M FFTs, the uncertainty should be S(f)/M1/2. Tara says to take M = 50. Possibly need to adjust formula to account for Welch overlap.

The above uncertainties are enough to estimate the statistical and systematic uncertainties on a fit to ϕc using the Nakagawa/Harry formula for a thin, lossy coating. By minimizing an appropriately weighted chi-squared function from 50 Hz to 500 Hz and then taking into account the above substrate and coating uncertainties, I find ϕc = (4.15 ± 0.03 stat ± 0.08 sys) × 10−4. More details will follow, and there may need to be some refinement (e.g., I still haven't dealt with the Welch overlap issue).

This has required adjusting the values of the Young modulus and Poisson ratio from their previous values (72.7 GPa and 0.167, respectively). I haven't checked these changes into the SVN.

I've tried harmonizing the Hong result (eq 94) with the Nakagawa/Harry formula, but the phi_tantala that I extract is about 9e-4, which is twice as high as previously reported values. I've spent some time hunting for a missing factor of two, but cannot find one.

  1406   Sun Feb 9 20:37:38 2014 EvanSummaryNoiseBudgetNoise budget fitting: result

Quote:

Quote:

In order to get the systematic uncertainty on ϕc, we need uncertainties in other parameters that enter the noise budget. Specifically:

  • Spot size, w. Currently using (182.0 ± 0.4) µm. The nominal comes from eq. 47 in Kogelnik and Li. This uncertainty can be propagated forward uncertainties in the following:
    • Cavity length, L. The spacer drawing (CTNLab/drawings/mechanical_drawings/dual_refcav/cavity_spacer_1.45inx1.5in.PDF) gives the length as (1.45 ± 0.01) inches.
    • Mirror radius, R. Nominal is 0.5 m; no idea about the uncertainty. It is not given in the test document. Currently using (500 ± 3) mm; i.e., an 0.5% uncertainty. This is the uncertainty claimed in the CVI catalog.
  • Coating thickness, d. Nominal is 4.5 µm; no idea about the uncertainty. Currently using (4.53 ± 0.07) µm from Tara's calculation (given in the reply to this elog post).
  • Substrate elastic modulus, Es. Using (73.1 ± 0.1) GPa as estimated from figure 29 in McSkimin 1953 (doi 10.1063/1.1721449), which is a (heavily cited) reference I found on the NIST ceramics database.
  • Substrate Poisson ratio, σs. Using 0.170 ± 0.004 as estimated from figures 29 and 30 in McSkimin (for Young and shear moduli, respectively) and then propagating the error forward to the Poisson ratio.
  • Cavity temperature. Should use 306(1) K.
  • Measurement uncertainty. Given an estimated PSD S(f) that is obtained by averaging M FFTs, the uncertainty should be S(f)/M1/2. Tara says to take M = 50. Possibly need to adjust formula to account for Welch overlap.

The above uncertainties are enough to estimate the statistical and systematic uncertainties on a fit to ϕc using the Nakagawa/Harry formula for a thin, lossy coating. By minimizing an appropriately weighted chi-squared function from 50 Hz to 500 Hz and then taking into account the above substrate and coating uncertainties, I find ϕc = (4.15 ± 0.03 stat ± 0.08 sys) × 10−4. More details will follow, and there may need to be some refinement (e.g., I still haven't dealt with the Welch overlap issue).

This has required adjusting the values of the Young modulus and Poisson ratio from their previous values (72.7 GPa and 0.167, respectively). I haven't checked these changes into the SVN.

On Tara's suggestion, I've done a fit to a coating loss angle with a power-law frequency dependence. The results are highly dependent on the band chosen for the fit (see attached plot).

For comparison, for a fit to a frequency-independent loss angle, the dependence on the band is much less prominent. For 50 Hz to 200 Hz, I get 4.12(3) × 10−4, and for 50 Hz to 700 Hz, I get 4.21(3) × 10−4.

Attachment 1: lossfit.pdf
lossfit.pdf
  1422   Mon May 5 10:37:24 2014 EvanSummaryDocumentationBayesian estimate of loss angles using Harry (2002)

Rather than using individual loss angles from Penn as the prior pdf, I've instead reanalyzed the data from Harry et al. (2002).

The ipynb for this is on the SVN in the noise budget folder.

Attachment 1: bayesian_harry.pdf
bayesian_harry.pdf bayesian_harry.pdf bayesian_harry.pdf
Attachment 2: marg_posts.pdf
marg_posts.pdf marg_posts.pdf
  1486   Wed Aug 27 03:21:53 2014 ranaSummaryopticoptimization for ETM with a-Si/SiO2 coatings

 I filled in more values for a-Si at 120 K into the wiki that Matt Abernathy set up. Then I ran the optimization code for Brownian noise only:

 aSi_120_Layers_60000.pdf

The above plot shows the comparison between the optimized aLIGO coating (silica:tantala at 300K) v. the a-Si coating at 120 K.

 aSi_R_60000.pdf

Then, finally, I compared the TO and Brownian noise of the two designs using the plotTO120.m script:

 aSi_120_TOnoise_60000.pdf

The dashed curves are silica:tantala and the solid lines are a-Si:silica. The Brownian noise improvement is a factor of ~6. A factor of ~1.6 comes from the temperature and the remaining factor of ~3.9 comes from the low loss and the lower number of layers.

I think this is not yet the global optimum, but just what I got with a couple hours of fmincon. On the next iteration, we should make sure that we minimize the sensitvity to coating thickness variations. As it turns out, there was no need to do the thermo optic cancellation since the thermo-elastic is so low and the thermo-refractive is below the Brownian almost at all frequencies.

  1531   Sun Nov 9 18:54:15 2014 EvanSummaryDocumentationHow to run the CTN experiment

A manual for running the CTN experiment is attached. I'll update and expand as needed.

Attachment 1: Manual.pdf
Manual.pdf
  1535   Tue Dec 16 14:06:35 2014 Aidan, Kate, EvanSummaryEnvironmentTask list

Some tasks not included on the list:

  • Temperature loops
    • Cavities
    • Can
    • PMC
  • Noise candidates:
    • Scatter
    • RFAM
    • PDH loop noise
Attachment 1: algaasTasks.jpg
algaasTasks.jpg
  1538   Wed Feb 4 17:29:29 2015 AidanSummaryDocumentationElectronics and power chassis connection summary

I spent this afternoon tracing out all the connections to and from all the chassis in the CTN lab.

We currently have 19 power supplies in use. Take a moment and think about that.

Attachment 1: labbook_1.pdf
labbook_1.pdf
Attachment 2: labbook_2.pdf
labbook_2.pdf
Attachment 3: labbook_3.pdf
labbook_3.pdf
Attachment 4: labbook_1a.pdf
labbook_1a.pdf
  1541   Mon Mar 30 14:29:25 2015 AidanSummaryEnvironmentCleaning underway

I've started removing a lot of the miscellaneous hardware from the lab (old pieces of Bosch framing, sheets of acrylic/plastic, etc). Some has gone into the ATF - we'll have to decide whether to keep it permanently or not. Right now, like Indiana Jones and the Last Crusade, I'm trying to see if there truly exists some space in this lab - or if it is a myth propagated though the ages. 

  1542   Tue Mar 31 11:43:37 2015 AidanSummaryComputersRebooted VME and added sitemap

Aidan

Aidan rebooted the Sun machine and VME. It took a while to get the EPICS channels to work again. The following seemed to work:

  1. Reboot Sun machine.
  2. Reboot the VME crate by depressing the reboot button on the top of the crate.
  3. Log into VME (psl1) at 10.0.0.2 from the Sun machine.
  4. Check the existence of various channels with dbpr "C3:PSL-RCAV_RCPID_SETPOINT"
  5. On the Sun, cd to /usr1/epics/psl/scripts and run "perl rcav_PID_2012_06_15.pl"
  6. Confirmed that PID values started updating on the Sun screen.

Aidan also added a sitemap (~/sitemap.adl), see attached image, for the CTN lab. Aidan added an alias to /home/controls/.bashrc

Aidan

The alias is:

alias sitemap="medm -x /home/controls/sitemap.adl"

Aidan

 

Attachment 1: Screenshot-sitemap.adl.png
Screenshot-sitemap.adl.png
  1545   Thu Apr 9 17:03:35 2015 AidanSummaryComputersAcromag ADC set up

I set up an Acromag slow controls based on the procedure that Keith wrote in T1400200. It's really pretty easy. It took an hour and 15 minutes from installing Ubuntu on a machine to having a functioning ADC channel from the Acromag unit. I haven't yet set up a DAC unit - this will require some tweaking of some of the EPICS parameters. Once I've done that I'll upload a complete procedure to the Wiki.

This is relatively promising for supporting/replacing VME slow channels.

  1546   Fri Apr 10 09:33:43 2015 EvanSummaryComputersAcromag ADC set up
Quote:

I set up an Acromag slow controls based on the procedure that Keith wrote in T1400200. It's really pretty easy. It took an hour and 15 minutes from installing Ubuntu on a machine to having a functioning ADC channel from the Acromag unit. I haven't yet set up a DAC unit - this will require some tweaking of some of the EPICS parameters. Once I've done that I'll upload a complete procedure to the Wiki.

This is relatively promising for supporting/replacing VME slow channels.

yesyes

  1547   Wed Apr 15 17:51:43 2015 AidanSummaryDAQAcromag XT1541 DAC working correctly - a little trick to configure

Aidan.

Success!

I configured the Acromag XT1541 DAC to run with EPICS. This was a touch trickier than the ADC as there is a subtlety with the channel configuration in the EPICS database. The bottom line is that now I can change the value in an EPICS channel and a multimeter attached to the unit will show a corresponding change in voltage.

The attached files (ioctest2.cmd and IOCTEST2.db) are used to access the first output channel, OUT00, on the unit. Now that I've got the thing working I can debug the calibration. Once that's sorted I'll summarize the set-up procedure on a Wiki page with glorious detail for future reference.

The following command line is used to open the modbus EPICS server.

${EPICS_MODULES}/modbus/bin/${EPICS_HOST_ARCH}/modbusApp ioctest2.cmd

The ioctest1 files are for the ADC unit.

Attachment 1: ioctest2.cmd
epicsEnvSet("IOC", "ioctest2")
epicsEnvSet("ARCH","linux-x86_64")
epicsEnvSet("TOP","${EPICS_MODULES}/modbus")
epicsEnvSet("MDBTOP","${EPICS_MODULES}/modbus")
dbLoadDatabase("$(MDBTOP)/dbd/modbus.dbd")
modbus_registerRecordDeviceDriver(pdbbase)
drvAsynIPPortConfigure("c3test2","10.0.1.41:502",0,0,1)
modbusInterposeConfig("c3test2",0,5000,0)
drvModbusAsynConfigure("DAC_Reg","c3test2",0,6,1,8,4,0,"Acromag")
dbLoadDatabase("./IOCTEST2.db")
... 2 more lines ...
Attachment 2: IOCTEST2.db
record(ao, "C3:ACROMAG_OUTPUT0")
{
	field(DTYP, "asynInt32")
	field(OUT, "@asynMask(DAC_Reg, 0, -16)MODBUS_DATA")
   	field(LINR,"NO CONVERSION")
	field(EGUL, "-163.84")
	field(EGUF, "163.835")
	field(HOPR, "163.835")
	field(LOPR, "-163.84")
	field(PREC, "2")
... 3 more lines ...
Attachment 3: ioctest1.cmd
epicsEnvSet("IOC", "ioctest1")
epicsEnvSet("ARCH","linux-x86_64")
epicsEnvSet("TOP","${EPICS_MODULES}/modbus")
epicsEnvSet("MDBTOP","${EPICS_MODULES}/modbus")
dbLoadDatabase("$(MDBTOP)/dbd/modbus.dbd")
modbus_registerRecordDeviceDriver(pdbbase)
drvAsynIPPortConfigure("c3test1","10.0.1.42:502",0,0,1)
modbusInterposeConfig("c3test1",0,5000,0)
drvModbusAsynConfigure("ADC_Reg","c3test1",0,4,0,8,4,32,"Acromag")
dbLoadDatabase("./IOCTEST1.db")
... 2 more lines ...
Attachment 4: IOCTEST1.db
record(ai, "C3:ACROMAG_INPUT0")
{
	field(SCAN, ".1 second")
	field(FLNK, "C3:ACROMAG_INPUT0_C.PROC")
	field(DTYP, "asynInt32")
	field(INP, "@asynMask(ADC_Reg 0 -16)MODBUS_DATA")
}
record(ai, "C3:ACROMAG_INPUT1")
{
	field(SCAN, ".1 second")
... 5 more lines ...
  1548   Thu Apr 16 13:31:32 2015 AidanSummaryDAQAcromag XT1541 DAC working correctly - a little trick to configure

https://nodus.ligo.caltech.edu:30889/ATFWiki/doku.php?id=main:resources:computing:acromag

 

Quote:

Aidan.

Success!

I configured the Acromag XT1541 DAC to run with EPICS. This was a touch trickier than the ADC as there is a subtlety with the channel configuration in the EPICS database. The bottom line is that now I can change the value in an EPICS channel and a multimeter attached to the unit will show a corresponding change in voltage.

The attached files (ioctest2.cmd and IOCTEST2.db) are used to access the first output channel, OUT00, on the unit. Now that I've got the thing working I can debug the calibration. Once that's sorted I'll summarize the set-up procedure on a Wiki page with glorious detail for future reference.

The following command line is used to open the modbus EPICS server.

${EPICS_MODULES}/modbus/bin/${EPICS_HOST_ARCH}/modbusApp ioctest2.cmd

The ioctest1 files are for the ADC unit.

 

  1549   Mon May 18 16:49:59 2015 AidanSummaryPMCRealigning PMC with rubbish input beam

I'm practicing the procedure of aligning the PMC. The input beam isn't great (I might be clipping a little on the Faraday). I'm mostly trying to get the hang of aligning into the cavity again as it's been a while.

My technique, so far, is to adjust the alignment, in yaw, of the third mirror before the cavity and then sweep the final mirror before the cavity slowly in yaw. I'm looking for flashes of transmission on the trans-camera. The goal is to keep tweaking until the order of the transitted mode is reduced to a few or zero. I should probably be sweeping the PZT at the same time.

I'm going to adjust the alignment through the FI tomorrow - just as soon as I get my hands on a couple more mirrors.

  1557   Thu Jul 30 14:55:14 2015 Aidan, AntonioSummaryEnvironmentBroken light replaced - general lab cleanliness

I had Facilities come and replace the dead light tube in the CTN lab antechamber. It's nice and bright in there right now.

We've noticed that the floor is pretty dusty - so we're implementing twice weekly mopping sessions starting tomorrow.

 

  1558   Sat Aug 1 09:24:58 2015 Aidan, AntonioSummaryBEATLasers locked to cavities - no beat - polarization? Transmission issues

Both lasers have been locked to the cavities for 24 hours. The slow control of the frequency is handed off to the PID loop. Antonio and I observed strange behaviour on the DC value of the cavity transmission.

As Evan had noted before, there are two polarizations that will resonate and they're about 3MHz apart (if I remember correctly). We can see these on the DC photodiodes on transmission (the ISS PD and the RF DC output). One peak is large and the other much smaller. However, when we have large transmission onto the RF photodiode we have small transmission onto the ISS PD and vice versa. It's likely we have a pick-off optic with strong polarization selectivity.

We couldn't find the beat yesterday or Thursday.

  1559   Sat Aug 1 09:29:25 2015 AidanSummaryEnvironmentCleaned computer table. Added Acromag

I cleaned the table with the computers and removed one of the monitors. I installed the Acromag units to the rack, powered them up and got them onto the network. They are:

XT1221 unit: 10.0.0.42

XT1521 unit: 10.0.0.41

I've pinged them successfully from other machines. I have an Ubuntu machine that I borrowed from the TCS lab to interface to them. I'll set up the EPICS control on Monday. In addition to adding the DC transmission channels to Acromag, we should be able to start migrating the PID controls away from the VME crates to these new units. 

 

 

  1560   Tue Aug 4 11:28:48 2015 Antonio, AidanSummaryEnvironmentDC transmission channels to Acromag and floor cleaning

Yesterday the DC transmission (ISS photodiode, North cavity) channel to acromag has been added and the viewer has been set on the Ubuntu machine. The channel has been also calibrated (against voltage externally injected). We are going to add other channels from the VME crates and work on the PDH board interface in order to remote control this unit.

Additionally the floor has been sweeped and mopped.

 

 

 

 

 

 

  1561   Tue Aug 4 14:15:14 2015 AidanSummaryComputersDAQD not running on FB2. FB2 trend disk is full

Like the title says ...

If I try running DAQD per https://nodus.ligo.caltech.edu:30889/ATFWiki/doku.php?id=main:experiments:psl:add_channel_for_daq_in_fb2

then it fails and the log file reveals that this is when it tries to write a GWF file to the trend folder. Manually trying to write anything to this location results in a "disk full" message.

The trusty df command yields the following.

[controls@fb2 frames]$ df
Filesystem           1K-blocks      Used Available Use% Mounted on
/dev/mapper/VolGroup00-LogVol00
                     232477448 125679200  94798608  58% /
/dev/sda1               101086     25961     69906  28% /boot
tmpfs                  1029664         0   1029664   0% /dev/shm
/dev/sdc1            240362656 205728852  34633804  86% /frames/full
/dev/sdd1            307663800 307663800         0 100% /frames/trend

 

 

  1562   Tue Aug 4 17:41:01 2015 Aidan, AntonioSummaryRefCavDAQ running with TRANS channels - strange peaks as we scan the laser frequency

We now have DC transmission channels in the frames. I'll post the details soon, but here's a plot that shows the transmission through the cavity on the RF_DC channel and the ACAV ISS DC as we slowly ramp the ACAV_SLOWOUT (temperature control) at a rate of 5E-5V/s.

Note the strange shape of the transmission peaks.

The wide view allows us to see the cavity transmit the upper and lower sidebands. Look carefully at the TRANS_RF_DC curve.

Attachment 1: ACAV_TRANS.png
ACAV_TRANS.png
Attachment 2: ACAV_TRANS_wide.png
ACAV_TRANS_wide.png
  1563   Tue Aug 4 18:37:21 2015 EvanSummaryRefCavDAQ running with TRANS channels - strange peaks as we scan the laser frequency
Quote:

We now have DC transmission channels in the frames. I'll post the details soon, but here's a plot that shows the transmission through the cavity on the RF_DC channel and the ACAV ISS DC as we slowly ramp the ACAV_SLOWOUT (temperature control) at a rate of 5E-5V/s.

Note the strange shape of the transmission peaks.

The wide view allows us to see the cavity transmit the upper and lower sidebands. Look carefully at the TRANS_RF_DC curve.

 
Bizarro.
 
Do you see the same thing if you unhook the PDs from the digital system and instead watch them on a scope?
 
Do you see the same thing if, instead of sweeping the laser temperature, you put a triangle wave on the laser PZT and scan across the resonances? My recollection is that the resonances (viewed on a scope) should be quite clean and lorentzian, apart from the issue of the two polarizations.
  1564   Wed Aug 5 10:18:54 2015 AidanSummaryComputersSummary of installation of Acromag units and control workstation

Summary of installation of new hardware:

 

  1. ACROMAG1 computer.
    1. This is a desktop running a new install of Ubuntu 12.
    2. It is on the CTN lab network and has been assigned the static IP address 10.0.0.33, with a Gateway 255.255.255.0
    3. We can’t seem to access the regular internet on this machine just yet.
    4. I’ve set up the standard CONTROLS account
    5. I’ve installed EPICS per https://dcc.ligo.org/LIGO-T1400200 that includes the MODBUS app that can access the ACROMAG units.
  2. ACROMAG units
    1. The XT1221 (ADC) and XT1541 (DAC) units are installed in the northwest rack
    2. They have static IP addresses:

i.XT1221: 10.0.0.42

ii.XT1541: 10.0.0.41

  1. They are getting a temporary +18V to power them.
  2. The XT1221 is attached to the router and the XT1541 is daisy-chained to the former.
  3. I need to check the averaging settings on the ADC unit to make sure they’re not overtaxing the system
  1. MODBUS/EPICS software
    1. I set up a MODBUS IOC per https://nodus.ligo.caltech.edu:30889/ATFWiki/doku.php?id=main:resources:computing:acromag
    2. The interface to the ACROMAG units is defined in acromag.cmd
    3. The input channels are defined in IOCTEST.DB and these report a value in counts

i.E.g. C3:ACROMAG_INPUT0

  1. We calibrated created three CALC channels that provide outputs calibrated in VOLTS. The calibrations are hardcoded into the database file.

i.C3:PSL-TRANS_ACAV_ISS_DC

ii.C3:PSL-TRANS_RCAV_ISS_DC

iii.C3:PSL-TRANS_RF_DC

  1. The two ISS channels are connected to their respective PDs
  2. The RF_DC output goes through an SR560 with a 30Hz pole and 50x gain on it to amplify the signal before the ADC.
  1565   Wed Aug 5 15:59:07 2015 Aidan, AntonioSummaryRefCavACAV transmission on RF photodiode

We swept the temperature of the ACAV laser and monitored the transmission through the cavity. In the attached image, the TEM00 modes (separated by 3GHz or 1 FSR) are located at C3:PSL-ACAV_SLOWOUT values of 0.5344V and 1.379V. There is a TEM20 mode at 1.52V.


 

Attachment 1: ACAV_FSr_full.png
ACAV_FSr_full.png
  1566   Thu Aug 6 23:49:10 2015 AntonioSummaryFSSpower transmitted through the cavities and control signal

The light transmitted from both the cavities has been monitored while the cavities where locked (Vacav = 1.369 V, Vrcav = 5.7909 V) and beats on RF photodiodes where visible. Power on the three photodiodes PD-Rcav (North cavity), PD-Acav (South cavity) and PD-RF decreases of about 10% from its maximum value on PD-RF and  about 5% on the others photodiodes periodically every ~7 minutes. 

I also notice that:

  1. The PD-Rcav trace is noisier than PD-Acav trace;
  2. The mean voltage values of the two photodiodes are way different:

PD-Rcav = ~60mV (+- 2/3%);

PD-Acav = ~170mV (+-2/3%);

       3. Enabling/disabling the boost switch on the FSS box does not give any improvement;

       4. Pressing the red botton (gain) on the FSS box neither;

 

In the same condition of locking the control signals of both PDH loop have been monitored too.

Here we can see that:

  1. the PDH-North loop is noisier;
  2. A step occurred in the cian trace (at the second division) without causing variation in the transmitted light;
  3. Enabling/disabling the boost switch on the FSS box does not give any improvement;
  4. Pressing the red botton (gain) on the FSS box neither aside from the fact that when the botton is kept pushed it slightly "cleans" the trace (not quantified);

It is worth also monitoring the error signal while the cavities are locked and when they are not (with a triangular wave applied at the laser PZT).

 

Control signal of PDH-North (yellow); Control signal of PDH-South (cian); PD-RF power (pink)
PD_RF
PD-RF
PD-Rcav (lower) , PD-Acav (upper)

 

Attachment 2: Windows_Phone_20150806_18_30_46_Pro.jpg
Windows_Phone_20150806_18_30_46_Pro.jpg
  1567   Mon Aug 10 17:49:41 2015 AntonioSummaryEOMerror-signal end EOM

On saturday a qualitative effect of the modulation produced by the EOM located in the PDH-north loop has been checked.

The goal was to have a look at the error signal of the PDH-north while the laser PZT was scanning frequecies around the two s-p TEM00 resonances. Because a that time I did not find the right error-signal connections on the FSS board (next elog will clarify where it is) I have demodulated the signal with an external mixer (and with a low pass filter) and monitored it. The picture shows the error-signal that we have with this setup: 

Caption

 

 

 

 

 

  1568   Mon Aug 10 18:44:01 2015 Antonio, AidanSummaryFSSFSS interface board

We figured out that when we sweep the laser PZT and we want to look at the error-signal, this can be monitored at the "mixer" output of the FSS interface board while the switch test/off/ramp is on ramp. However we need to understand what this switch does when it is OFF.

 

 

 

 

 

 

Attachment 2: D0902049-v1(front_panel).pdf
D0902049-v1(front_panel).pdf
  1569   Tue Aug 11 16:49:55 2015 Aidan, Rich, AntonioSummaryBEATDiagnosing the terrible beat signal. RCAV HV supply not showing current drawn

The beat signal looks awful. It has some amplitude modulation at 6.75MHz and looks like it has some strange saturation effects going on. This is too much noise for the PLL to lock to.

We thought, for a minute, that the reason for this may be related to one of the HV supplies for the RCAV locking. The needles on the front of the positive supply unit +150V and 0mA current drawn. The other 3 HV supplies in use all show around 20-25mA current draw when used with the TTFSS boards.

We popped the top of the RCAV TTFSS box on the table and looked at the TP4 output on the HV/Interface board (this looks at the signal coming out of the high voltage amplifier that feeds the EOM, but reduced a voltage divider to 1/10th the value). It was freely swinging between +/-4V, so the HV amplifier seems to be happily getting both +ve and -ve voltages. There might be a problem with the needle on the HV supply.

 

Attachment 1: photo_1.JPG
photo_1.JPG
Attachment 2: photo_2.JPG
photo_2.JPG
  1570   Thu Aug 13 09:07:04 2015 Aidan, Rich, AntonioSummaryBEATPLL locked on beat note - TTFSS boxes were being used incorrectly previously

We investigated the way we were locking the PDH loops using the TTFSS boxes. Here's what we previously did:

  • With the loop open and the TTFSS interface set to LOCAL & TEST, we adjusted the temperature of the laser until we were close to TEM00 resonance.
  • Then we set the first switch from LOCAL to REMOTE. 
  • This locked the loop.

We realized, yesterday, that this wasn't the correct way to lock the loop as box now expected the gain settings for the loop to be set remotely (and we were providing none of that to the unit). Still, the default gain in REMOTE was enough to provide a stable lock and we didn't understand exactly how that box worked (which is obvious in retrospect). So, yesterday, we pored over the schematics for the TTFSS boxes (Rich is drawing a very nice block diagram to show the loop structure), and realized our error. The correct way to lock is the following:

  • Set to LOCAL & TEST
  • Adjust the temperature so we were close to TEM00 resonance.
  • Scan through the two polarization TEM00 eigenmodes (separated by ~1MHz)
  • We can sit (a) outside the mode with big transmission (big mode), (b) in between the two modes or (c) outside the small mode. 
  • We sit outside the big mode and then switch TEST to OFF to turn the loop on (I have been cursing about the naming conventions on these boxes for the last two days).
  • This locks the loop.
  • With the gains set locally.

From here, we were able to play with the common mode gain settings reduce the noise of the beat note between the lasers. And we were able to lock the PLL. The main evidence for the latter is the fact that we can change the DC value of the control signal in the PLL by varying the carrier frequency of the Marconi.

  1571   Fri Aug 14 19:10:39 2015 AidanSummaryTempCtrlDropped lab temperature set point from 72F to 68F

At 7:07PM

 

  1572   Mon Aug 17 17:02:41 2015 AidanSummaryComputersAcromag binary outputs working - still can't control TTFSS box though

I managed to figure out the modbusDrv configuration settings to get the binary output of the Acromag working. I've updated the Wiki page to reflect this. I've wired the XT1541 DAC, BIO Acromag unit to the T1EN and T2EN channels on the TTFSS box but I still can't get remote control of it yet for some reason. When the PDH loop is closed and I switch the TTFSS box to REMOTE, the loop stays closed regardless of what I do to the binary outputs in EPICS.

https://nodus.ligo.caltech.edu:30889/ATFWiki/doku.php?id=main:resources:computing:acromag

  1573   Wed Aug 19 10:34:33 2015 AidanSummaryDAQBinary IO control of TTFSS box working - can lock PDH remotely

I've worked out how to get the binary IO to work with the TTFSS box so that we can activate switches in that unit. It wasn't working in the setup yesterday because of physics. Actually - there is a 10K pulldown resistor in the Acromag unit that attaches the output to ground. The actual circuit looks like this:

 

VCC (5V) --- (4.99K) --- T1EN -----|-----DIO0 ----(6.2V if DOUT set to 1)---- (10K) ------ | GND

.........   TTFSS...................................| ..............................ACROMAG ...............................................|

T1EN is measured by the switch-chip (SN74HCT157D, chip U3 in D040423) to determine whether it should be open or closed. We need to bring T1EN below 0.8V to get the TTL logic to work.

If DOUT is set to 1, then DIO0 and T1EN become the excitation voltage, 6.2V, and the switch circuit reads high. If DOUT is set to 0, the excitation voltage is removed and we just end up with a voltage divider and around 3.33V at T1EN - which does not register as low.

We can get around this by adding a smaller resistor, say 810 Ohms, in parallel to the 10K, to lower the effective resistance of the pull-down resistor to 750 Ohms. The maximum current the Acromag unit will have to supply is 6.2V/750 Ohms = 8.4mA.

So that's what I did. Now, when I switch DOUT to 1, I see 6.2V at T1EN and when I switch DOUT to 0, I see 0.669V at T1EN. The TTFSS box registers these as two different states and I can lock and unlock the PDH loop from EPICS.

  1574   Wed Aug 19 13:53:45 2015 AidanSummaryDAQDAC calibration for Acromag slow units

Volts (requested) to Counts

C = a1*V_req + a2

a1 = 3003.5 counts/Volt

a2 = -20.8285 counts

Entered counts (in +ve and -ve amounts) and measured the resulting voltage across the output. I'm not sure how stable this calibration is.

Attachment 1: Photo_on_8-19-15_at_1.51.33_PM.jpg
Photo_on_8-19-15_at_1.51.33_PM.jpg
  1575   Thu Aug 20 13:55:22 2015 AntonioSummaryFSSOLTF Transfer functions of the North loop

In order to have a better understanding of the FSS electronic boards we decided to take

open loop transfer functions (OLTF) of the FSS North cavity loop. In the same time we took 

the opportunity to verify that the broadband EOM is working. However this measurement will

be done again for a better quantitative analysis of the loop performances.

 

OLTF measured in the common path shows the shape due to the PZT and due to the EOM

paths as expected.The UGF has been pushed up to 170kHz. In the picture below we see the

traces corrisponding to the two different gain settings.

 

Connections:

  

Network Analyzer                           FSS

           RF          ----------->        EXC (Common)

           R            ----------->        OUT2 (Common)

           A            ----------->        OUT1 (Common)

 

These measurements have been done with the Switch Exc on located on FSS interface board

and with the following gain settings:

 

  Fine Common Fast Offset

Initial settings

580 312 240 506
High gain 950 950 950 506

 

 

 

OLTF measured in the FAST path have the connections applied at the Fast and the switch EXC on

the front panel turned on. This measurement has been taken with "high gain settings" of the previous

measurement.In the following pictures can be seen the previous OLTF too (green/red).

 

Connections:

 

Network Analyzer                           FSS

           RF          ----------->        EXC (Fast)   (front panel switch Exc ON)

           R            ----------->        OUT2 (Fast)

           A            ----------->        OUT1 (Fast)

 

 

PLEASE note these are not the total OLTF (relative to their path) because is still missing the transfer function

between OUT1 and OUT2 in both the paths Common and Fast.

 

I would like to model this loop, I need to figure out the best way to do it.

ELOG V3.1.3-