40m QIL Cryo_Lab CTN SUS_Lab TCS_Lab OMC_Lab CRIME_Lab FEA ENG_Labs OptContFac Mariner WBEEShop
  ATF eLog, Page 2 of 56  Not logged in ELOG logo
ID Date Authorup Type Category Subject
  270   Tue Aug 18 08:50:56 2009 AidanComputingDAQOverflows on DAQ channels

Spotted a ton of accumulating overflows on the DAQ - specifically the output for DMass's PZT on his PMC.

It looks, from the EPICS screen, like that channel is trying to output a lower number of counts (~ -90000) than the DAQ can reasonably drive. (+/- 20V = 64K counts). We should try and set upper and lower limits on the channels. I'm not sure why this isn't set automatically in the RCG.

Also made a note on the ATF wiki ...

Seems to be a non-fatal error though.

Attachment 1: Screenshot.png
Screenshot.png
Attachment 2: Screenshot-3.png
Screenshot-3.png
  273   Tue Aug 18 17:49:52 2009 AidanLaserFiberMZ beam overlap after tweaking alignment: better @ 9% but still low

I improved the alignment into the recombination beam splitter using the TV camera to overlap the beams in the near and far fields of the splitter - essentially making them as colinear as possible. As  fixed the position of PD2 so that it was at the focus of the focussing lens and optimized the position of the beams on the two photodetectors.

I remeasured the AC and DC outputs from the photodetectors. The results are below:

 

Photodetector outputs - voltages
  PD1 PD2
Fiber DC 98.5 +/- 0.5mV 69.2 +/- 0.3mV
Reference DC 65.3 +/- 0.2mV 113.5 +/- 1mV
AC @ 80MHz (dBm) +5.5 +/- 0.1 dBm +6.5 +/- 0.05dBm
AC @ 80MHz (volts rms) 421mV 473mV

 

Photodetector output - converted to power
  PD1 PD2
Fiber DC 131 uW 92 uW
Reference DC 87 uW 151 uW
AC @ 80MHz (amplitude) 20 uW 22 uW
overlap 9.3% 9.3%

overlap = AC@80MHz/(2*sqrt[FiberDC*RefDC])

This leaves the beam size as the culprit for the relatively poor overlap. The reference beam size is noticably bigger than the fiber beam size on the CCD. I'll see if I can improve this ...

 

  274   Wed Aug 19 09:42:38 2009 AidanComputingDAQOverflows: PMC PZT output in the billions of counts ... burtrestore issue?

Came in and found the PMC_PZT output trying to deliver an obscenely large number of counts to the DAC.

Not really sure where the issue is coming from - looks like the filters in the attached module. Anyway, just set a limit on what that module can output for the time being. Will leave this one for DMass to figure out.

Frank was talking about rebooting the frame builder last night. That shouldn't have affected the front end though. But if we do want to reboot the front end will all our settings be saved and restored automatically?

 

Attachment 1: Screenshot.png
Screenshot.png
  279   Fri Aug 21 18:44:33 2009 AidanLab InfrastructureGeneralMarconi Memory Setting 103 - 80.0001MHz

Added another setting to the memory of the Marconi - #103.

Same as memory setting #102 but with the added 100Hz signal - can be used to easily calibrate the output voltages of the mixers corresponding -pi/2 and +pi/2

See http://131.215.115.52:8080/AdhikariLab/248 for the other settings.

  281   Sat Aug 22 21:04:28 2009 AidanLaserFiberClosed Loop on Single Pass again - strange saturation at low freq.

After improving the overlap between the two beams on the PDs and increasing the power incident on each PD, I closed the loop again.

The loop looks like this:

PD1---Demodulated_AC---SR560---Marconi_FM_IN---AOM---PD1

I played around with the SR560 settings and the Marconi settings. The spectra of phase noise are not calibrated although the saturation in the open loop signal indicates the number of counts that correspond to pi radians. The DAQ noise floor is just below 10^-2 counts.

I played around with the phase of the LO in the out of loop path so that the demoudlated AC signal on it was centered around 0V.

At this stage I need to add a whitening filter and amplify the signal before putting it into the DAQ. I think I'll try closing the loop with the DAQ output as well. Also, I don't understand why the out-of-loop signal appears to be saturating at lower frequencies when the loop is closed.

Note: FM Devn is the deviation in the output frequency of the Marconi per unit volt (rms) [Units: Hz/V]

Attachment 1: 2009-08-22_uncal_phase_matched_SR560.pdf
2009-08-22_uncal_phase_matched_SR560.pdf
  282   Sun Aug 23 12:37:54 2009 AidanLaserFiberLow Freq Acoustic Noise source [50-200Hz] identified

 

When testing the performance of a loop in the range between 50 and 200Hz, I recommend using The Dark Knight Soundtrack - Track 11 as an acoustic noise source.

 

Attachment 1: 2009-08-23_low_freq_acoustic_noise.pdf
2009-08-23_low_freq_acoustic_noise.pdf
  283   Sun Aug 23 12:41:18 2009 AidanMiscGeneralLower lab temperature 3 degrees. Too warm in here.

Changed the set point on the temperature in the lab to 61F from 64F.

  284   Sun Aug 23 18:21:32 2009 AidanLaserDAQWhitening Filter for Fiber signals into DAQ
Attachment 1: FILT-AB1.zip
  309   Tue Sep 15 13:30:02 2009 AidanElectronicsFiberWhitening Filter transfer function

The whitening filter added to the FS out of loop path.

Pole  ~73Hz

Zero ~0.162Hz

Gain ~ 0.9x

 

 

Attachment 1: WHITE001.MAT
Attachment 2: WHITE001.ASC
Input File: white001.78d
Measure Group: Swept Sine
Measurement: Freq. Resp.
Num of extracted Points: 160
Start Freq.: 10 kHz
Stop  Freq.: 10 mHz
Sweep Type: Log
Auto Level: Off
Amplitude:  10.0 mV
Source Ramping: Off
... 168 more lines ...
Attachment 3: Whitening_FIlter_-_SR785.pdf
Whitening_FIlter_-_SR785.pdf
Attachment 4: 2009-09-15_Whitening_Filter_Fit.png
2009-09-15_Whitening_Filter_Fit.png
  311   Tue Sep 15 16:45:04 2009 AidanElectronicsFiberEffect of whitening filter on fiber stabilization out of loop PD

The additional passive whitening filter (zero at 0.34Hz and pole at 83.7Hz) and the 500x gain added to the out of loop photodetector has greatly whitened the out of loop measurement of fiber phase noise. The attached plot shows the raw (counts not radians) measurements of the whitened and dewhitened spectra of the out of loop stabilized fiber noise as well as the noise with no whitening filter attached.

Note that the gain of the whitening filter is ~1x at DC which is why all three spectra have the same level at DC.

Attachment 1: 2009-09-15_Whitening_dewhitening.pdf
2009-09-15_Whitening_dewhitening.pdf
  314   Wed Sep 16 06:52:57 2009 AidanElectronicsFiberWhitening Filter transfer function

I don' trust the measured and fitted transfer function because it's a little noisy and they also look railed at very low frequencies. I measured the values of the resistors and capacitor in the filter and calculated the following values:

zero at 0.342Hz, pole at 83.8Hz, gain at DC = 0.00199. These are the values I will use for the time being.

 

Quote:

The whitening filter added to the FS out of loop path.

Pole  ~73Hz

Zero ~0.162Hz

Gain ~ 0.9x

 

 

 

  315   Wed Sep 16 09:49:51 2009 AidanElectronicsFiberOpen and closed loop fiber noise with wnd without whitening filter

Effect of whitening filter is shown in attached plot. With the loop closed and no filter we lose dynamic range at high frequencies. It looks like the out of loop measurement of fiber noise is limited partly by the DAQ noise, whereas with the range is increased to accomodate all the frequencies from the fiber noise.

I'm not sure why the DAQ noise is slightly lower than the limit of the no WF closed loop spectra. Maybe a calibration issue?

Going to increase the loop gain, try to suppress the fiber noise some more and see if I can still see the signal without being limited by DAQ noise.

Attachment 1: 2009-09-16_whitening_filter_open_closed_loop.pdf
2009-09-16_whitening_filter_open_closed_loop.pdf
Attachment 2: 2009-09-16_whitening_open_closed_2x50KHz.jpg
2009-09-16_whitening_open_closed_2x50KHz.jpg
  321   Wed Sep 16 17:15:59 2009 AidanComputingFiberWhitening filter and noise spectra

As before but with electronics noise added.

 

Attachment 1: 2009-09-16_whitening_filter_open_closed_loop_electronics.pdf
2009-09-16_whitening_filter_open_closed_loop_electronics.pdf
  322   Thu Sep 17 08:22:16 2009 AidanLaserGeneralChiller warning - DI cartridge needs changing

The 35W laser chiller was beeping this morning when I came in. The status message was 'Chang DI'.

This pertains to the DI cartridge. I checked the Puralite sensor on the back and it's red so we'll need to change the DI cartridge.

 

 

Attachment 1: Screenshot.png
Screenshot.png
  323   Thu Sep 17 09:48:19 2009 AidanLaserFiberClosed loop with DAQ. Success.

I closed the fiber noise stabilization loop through the DAQ. The error signal from PD1 goes to a whitening filter goes through an SR560, goes to the DAQ, is dewhitened and amplified and then output to analog and fed into the FM input on the Marconi. When the loop closes there is a clear suppression of the noise from the PD.

I also tee'd off the PD signals for both in and out of loop before the whitening filter and fed these into the DAQ as a sanity check.

Also: the loop must be opened by turning off the option for external modulation on the Marconi (which drives the AOM) rather than just turning off the error signal in the DAQ. The reason for this is that the output noise from a DAQ channel appears to be sufficient to create noise on the output from the AOM. Which doesn't quite make sense because that noise isn't present when the loop is actually closed. Will look into this.

The effect of closing the loop can be seen in the following time series from the whitened channels, the dewhitened versions of these and the unwhitened witness channels from the PDs. The loop was closed around t = -7.0s.

Attachment 1: Screenshot-1.png
Screenshot-1.png
  324   Thu Sep 17 09:53:06 2009 AidanLab InfrastructurePurchasesLighting for bench

I purchased some strip lighting from McMaster Carr to illuminate the bench opposite the computer. They fit quite nicely but I'm a little disappointed with the brightness.

 

Attachment 1: photo.jpg
photo.jpg
  327   Thu Sep 17 16:14:36 2009 AidanLaserFiberYesterday's spectra with shot noise included

Here is yesterday's plot of the suppressed whitened-dewhitened fiber noise with the approximate shot noise added. The measured noise spectra is within a factor of 2 of this level. I'll try pushing it a little harder in addition to getting the locking with the DAQ working.

Attachment 1: 2009-09-16_whitening_filter_open_closed_loop_electronics_shot.pdf
2009-09-16_whitening_filter_open_closed_loop_electronics_shot.pdf
  328   Thu Sep 17 18:22:05 2009 AidanComputingFiberLoop closed with DAQ. Script running overnight

 

Am running the following script overnight to increase the loop gain for the fiber loop every 4 hours.

The process is gain_change.sh

 

#!/bin/csh
# - - - - - - - - - - - - - - - - - - - -
# scan the FS loop gain up from 100 to 1000 in 4 hour increments
# 17th Sept 2009


   sleep 5
   ezcawrite "C2:ATF-GENERIC_GENOUT8_GAIN" "100"
   sleep 14400
   ezcawrite "C2:ATF-GENERIC_GENOUT8_GAIN" "330"
   sleep 14400
   ezcawrite "C2:ATF-GENERIC_GENOUT8_GAIN" "1000"

 

 

 

  338   Fri Sep 18 16:07:15 2009 AidanLaserFiber(nearly) Calibrated noise spectra through DAQ

I've closed the loop through the DAQ and have measured the phase noise and frequency noise as shown in the attached plot. There's a lot more that needs to be taken care of when the loop is closed in this way and it doesn't help that the input SR560s introduce a DC offset every time I change the gain on them. There is also possibly a factor of Pi that I need to include in the phase and freq plots. Additionally I need to determine what the output DAC noise level is.

However, the noise is clearly reduced in this way, I'm not limited by DAC noise (but might be limited by ADC noise) but am again limited by electronics noise. Anyway, part of the purpose of this was to just close the loop with the DAQ and determine if I could stabilize the noise that way. I'm still getting to grips with all of the details of this. For instance, I think, and Frank agrees, that I might need a whitening filter on the output signal from the DAQ and then need to dewhiten with an analogue filter outside in the real world.

A schematic of the current loop is shown below:

Attachment 1: 2009-09-18_closed_loo_DAQ.pdf
2009-09-18_closed_loo_DAQ.pdf
Attachment 2: 00001.png
00001.png
  344   Tue Sep 22 14:45:51 2009 AidanElectronicsFiberDAQ transfer function - phase delay bit me as well

Like others before me I have been bitten by the phase delay in the DAQ. The transfer function of the FS DAQ setup, as illustrated below, was recorded and is attached:

VIn --> whitening filter --> SR560 gain --> (to DAQ)

       [ input anti-whitening and gain --> output anti-dewhitening and gain ] -->

            (from DAQ) --> dewhitening filter ---> VOut

An analysis of the slope of the phase indicates that the time delay through the DAQ is around 0.4ms. Ho hum.

 Settings:

  • Whitening Filter: [zero:  0.342Hz, pole: 85.77Hz, gain @ DC = 1/500]  **
  • input SR560 gain = 100x
  • input anti-whitening filter: [pole: 0.342 Hz, zero: 85.77Hz, gain @ DC = 500x]
  • input module gain = 1/100
  • output anti-dewhitening filter: [zero: 2.19 Hz, pole: 111.85Hz, gain @ DC = 1x]
  • output gain = 10x
  • dewhitening filter: [pole: 2.19Hz, zero: 111.85Hz, gain @ DC = 1/2] **

* (Volts per count on the output is half as much as on the input; therefore an additional factor of 1/2 through the DAQ)

** (zero and pole values are calculated from measured values of resistors and capacitors)

Attachment 1: DAQTFFS1.zip
Attachment 2: 2009-09-22_DAQ_TF.png
2009-09-22_DAQ_TF.png
Attachment 3: 2009-09-22_DAQ_transfer_function.pdf
2009-09-22_DAQ_transfer_function.pdf
  346   Wed Sep 23 19:29:38 2009 AidanElectronicsFiberVCO and mixer transfer function - saturated.

VCO1E3HZ.MAT - transfer function, Vin = 500mV amplitude, 0mV offset, FMDevn = 1kHz

VCO1E5HZ.MAT - transfer function, Vin = 500mV amplitude, 0mV offset, FMDevn = 100kHz

VCO1E5LO.MAT - transfer function, Vin = 50mV amplitude, 0mV offset, FMDevn = 100kHz

 

The saturation at lower frequencies is somehow related to the maximum output of the mixer.

Attachment 1: VCO1E5HZ.zip
Attachment 2: 2009-09-23_VCO_transfer_function.png
2009-09-23_VCO_transfer_function.png
Attachment 3: 2009-09-23_VCO_transfer_function.pdf
2009-09-23_VCO_transfer_function.pdf
  356   Mon Sep 28 20:40:53 2009 AidanMiscstuff happensPlease do not leave the permanent marker by the whiteboard ...
  363   Tue Oct 6 23:37:06 2009 AidanLab InfrastructureGeneralEnclosure and cable tray installation

The cable trays were installed overhead last Thursday and Friday. Tables were covered with large plastic sheets. The installation created a little bit of mess but repeated lab cleaning by Frank et al has dramatically improved the cleanliness of the lab. (I'm sure he'll have more to say on that).

Construction of the Bosch enclosure was begun on Friday and completed on Monday (except for the acrylic sheets for the ceiling). The shelf has been installed and is bolted to the structure on either end.

At last the cables are off the floor and the electronic equipment is off the optical table.  BTW, we may need to lay some BNC cables across the floor until we get ones which are long enough for the cable trays.

Thanks to all for help installing all this stuff. And for the help cleaning the lab. Much appreciated.

 

 Check out the before and after ...

Attachment 1: before.jpg
before.jpg
Attachment 2: after.jpg
after.jpg
  364   Wed Oct 7 00:05:17 2009 AidanMiscFiberIntraaction 70 MHz AOM - photos

Here are some photos of the inside of the 70MHz Intraaction AOM that we're using temporarily in the FS experiment. I'm not sure what the crazy red glue is all about.

Attachment 1: aom1.jpg
aom1.jpg
Attachment 2: aom2.jpg
aom2.jpg
Attachment 3: aom3.jpg
aom3.jpg
Attachment 4: aom4.jpg
aom4.jpg
Attachment 5: aom5.jpg
aom5.jpg
  365   Wed Oct 7 00:51:35 2009 AidanLaserFiberAOM installed and running on far side of fiber

I've installed an AOM on the far side of the fibre to shift the frequency of the transmitted beam by 70MHz. We're driving it with a +2dBm signal which is being amplified by ~33dB by a ZHL-1-2W amplifier from MiniCircuits. We're getting around 12% of the total optical power in the first order beam. It is also extremely elliptical.

In the image below: the output of the fiber coupler the beam is directed into an AOM where the first order beam is around 12% of the input. The first order beam passes through an iris which block the zeroth order beam. It goes through an f = 150mm lens to produce a waist which can be used with a flat for a retro-reflection. Unfortunately, when the beam is viewed after this lens with the beam scan, it is obvious that it is extremely elliptical. The retro-reflector, a flat partial reflector (R = 85% at normal incidence), is placed near the waist of the horizontal, which is about 305 um. The waist in the vertical appears to be at the lens and around 700-750um radius.

The beam, originally p-polarized, goes through a QWP which converts it to circular, it is "retro-reflected" and  goes back through the QWP to become s-polarized. It works it way back to the AOM where there is another zeroth order transmission and a first order transmission. The zeroth order is the faint beam that is dumped on the fiber coupler. The first order transmission is co-linear with the original beam and couples back in to the fiber.

Back at the input end of the fiber the double passed beam is extracted using a polarizing beam splitter. Rana and I optimized the transmission with the retro-reflector in it's current position but it's quite weak (we haven't yet determined how much power makes it back).

Attachment 1: 00001.jpg
00001.jpg
  366   Wed Oct 7 01:14:26 2009 AidanLaserFiberAOM installed and running on far side of fiber

Quote:

I've installed an AOM on the far side of the fibre to shift the frequency of the transmitted beam by 70MHz. We're driving it with a +2dBm signal which is being amplified by ~33dB by a ZHL-1-2W amplifier from MiniCircuits. We're getting around 12% of the total optical power in the first order beam. It is also extremely elliptical.

In the image below: the output of the fiber coupler the beam is directed into an AOM where the first order beam is around 12% of the input. The first order beam passes through an iris which block the zeroth order beam. It goes through an f = 150mm lens to produce a waist which can be used with a flat for a retro-reflection. Unfortunately, when the beam is viewed after this lens with the beam scan, it is obvious that it is extremely elliptical. The retro-reflector, a flat partial reflector (R = 85% at normal incidence), is placed near the waist of the horizontal, which is about 305 um. The waist in the vertical appears to be at the lens and around 700-750um radius.

The beam, originally p-polarized, goes through a QWP which converts it to circular, it is "retro-reflected" and  goes back through the QWP to become s-polarized. It works it way back to the AOM where there is another zeroth order transmission and a first order transmission. The zeroth order is the faint beam that is dumped on the fiber coupler. The first order transmission is co-linear with the original beam and couples back in to the fiber.

Back at the input end of the fiber the double passed beam is extracted using a polarizing beam splitter. Rana and I optimized the transmission with the retro-reflector in it's current position but it's quite weak (we haven't yet determined how much power makes it back).

 The double passed beam is extracted using a PBS and is incident on a photo-detector, as shown in this diagram. There is a, relatively, large amount of scatter (about 20-30% of PD signal) which appears to be coming from the AOM. Right now we're not doing anything about it, because it is a different frequency and polarization to the double-passed light, but we should look into clearing it up in the near future.

(I haven't shown the zeroth order beam from the AOM in this picture).

Attachment 1: PBS.jpg
PBS.jpg
  367   Wed Oct 7 02:47:41 2009 AidanMiscstuff happensDigital camera is in my office
  370   Wed Oct 7 22:00:43 2009 AidanLaserFiberBeam profile of double-passed transmission from fiber.

This beam was picked off by the PBS, as shown in the previous elog. It looks extremely Gaussian. It goes through a waist around 1.5m from the PBS. This is at about 2.4m from the PBS.

I measured the beam at half a dozen places and I'm trying to fit to this, but I need to interpret exactly what the DataRay beamsize is defined as.

 

Attachment 1: 2009-10-07_double_passed_fiber.PNG
2009-10-07_double_passed_fiber.PNG
  371   Wed Oct 7 22:04:49 2009 AidanLaserFiberFS layout

Quote:

I took a bunch of pics of the FS experiment layout and stitched them together to show the layout.  I saved the intermediate gimp files so this can be editted if anything is missing or not right.

 

EDIT: added label for the faraday isolator to the diagram

 

 

Can you paint the fiber in a differnt colour to the beam; I suggest blue. Please use PNG for pictures.

  375   Sat Oct 10 00:44:36 2009 AidanLaserFiberMZ phase/frequency noise

Since the phase noise in the reference arm in the fiber stabilization experiment will be imposed on the laser I thought I'd check out what the phase noise of a bog-standard, unlocked MZ with Newport Ultima mounts and no acoustic shielding was.

Attached is a measurement of the phase and frequency noise of just such an interferometer. The rms frequency noise below 20Hz is around 17mHz, well below the 75mHz limit set by Bram (to get 1pm motion in the arms), but the noise at higher frequencies contributes a lot more. The rms below 256Hz is around 580mHz.

I'm going to try again with an acoustic enclosure around the MZ. There's no reason we should be including acoustic noise and air currents in our measurements.

Attachment 1: MZ_phase_noise.pdf
MZ_phase_noise.pdf
  424   Fri Nov 6 19:46:57 2009 AidanLaserFiberVCO phase noise requirements for Fiber Stabilization

 I calculated the maximum phase noise allowable from the VCO. This noise will be suppressed by the same amount as the open loop fiber noise (suppression ~ loop gain) so it will not be dominant provided it is less than the open loop fiber noise floor. I took the open loop phase noise accumulated in a 4km fiber from Hanford and set the maximum allowable VCO phase noise to be 10x smaller.

Thus an acceptable level VCO phase noise is ~

0.8 radians/Hz^{1/2} @ 1Hz with an ~ 1/f dependence. See attached plots.

Attachment 1: VCO_requirements.pdf
VCO_requirements.pdf
Attachment 2: VCO_requirements.jpg
VCO_requirements.jpg
  426   Sun Nov 8 13:14:43 2009 AidanLaserFiberCorrection: VCO phase noise requirements for Fiber Stabilization

Quote:

 I calculated the maximum phase noise allowable from the VCO. This noise will be suppressed by the same amount as the open loop fiber noise (suppression ~ loop gain) so it will not be dominant provided it is less than the open loop fiber noise floor. I took the open loop phase noise accumulated in a 4km fiber from Hanford and set the maximum allowable VCO phase noise to be 10x smaller.

Thus an acceptable level VCO phase noise is ~

0.8 radians/Hz^{1/2} @ 1Hz with an ~ 1/f dependence. See attached plots.

I had a bit more of a think about the VCO phase noise requirements. The attached plots are the phase noise requirements for the oscillators at points A and B in the attached diagram. However the most restrictive requirement for the phase noise comes from the local oscillator at point C in the attached diagram. This will not be suppressed by the loop gain and therefore can easily become the limiting noise. Still need to work out the requirements for the phase noise at this point.

 

 

Attachment 1: FS_control_loop.png
FS_control_loop.png
Attachment 2: FS_control_loop.pdf
FS_control_loop.pdf
  428   Sun Nov 8 14:23:46 2009 AidanLaserFiberVCO as LO phase noise requirements for FS

Quote:

Quote:

 I calculated the maximum phase noise allowable from the VCO. This noise will be suppressed by the same amount as the open loop fiber noise (suppression ~ loop gain) so it will not be dominant provided it is less than the open loop fiber noise floor. I took the open loop phase noise accumulated in a 4km fiber from Hanford and set the maximum allowable VCO phase noise to be 10x smaller.

Thus an acceptable level VCO phase noise is ~

0.8 radians/Hz^{1/2} @ 1Hz with an ~ 1/f dependence. See attached plots.

I had a bit more of a think about the VCO phase noise requirements. The attached plots are the phase noise requirements for the oscillators at points A and B in the attached diagram. However the most restrictive requirement for the phase noise comes from the local oscillator at point C in the attached diagram. This will not be suppressed by the loop gain and therefore can easily become the limiting noise. Still need to work out the requirements for the phase noise at this point.

 

 

Worked out a rough phase/frequency noise floor requirement for the VCO. I assumed that the frequency noise was white (which is a gross oversimplification) and worked out that a level of around 0.1mHz/Sqrt(Hz) would give around 7.5mHz rms noise (10x less than requirement for fiber stabilization) below 5kHz. The phase noise spectrum is simply this times 1/f.

The spectra are attached. The next step is to take a slightly more realistic phase noise spectra - say from the VCOs at LHO (http://ilog.ligo-wa.caltech.edu:7285/advligo/LowNoiseVco) and determine if they yield low enough frequency noise.

 

 

Attachment 1: VCO_PN_requirement.pdf
VCO_PN_requirement.pdf
Attachment 2: VCO_PN_requirement.png
VCO_PN_requirement.png
  429   Mon Nov 9 10:36:04 2009 AidanLaserFiberVCO as LO phase noise requirements for FS

 Added Sigg's model phase noise numbers (had to interpolate down to 1Hz from 10Hz) and Mott's Marconi measurements to the VCO phase noise plot.

Reminder: the maximum frequency noise, for ALS and hence fiber stabilization, is 75mHz rms (this will saturate ESDs). The target for the VCO is 10x less than this = 7.5mHz.

Looks like the Marconi is not quite good enough, while the LHO SIgg Oscillator is below the target RMS value of 75mHz. Ideally the RMS noise of the oscillator would be a factor of 10x smaller than the maximum allowed noise but a factor of 7 or so should be sufficient.

I need to get the original data from the Marconi PN measurement. I had to estimate the plot from Mott's original measurement here ...

Attachment 1: VCO_phase_noise_requirements.png
VCO_phase_noise_requirements.png
Attachment 2: VCO_phase_noise_requirements.pdf
VCO_phase_noise_requirements.pdf
Attachment 3: VCO_freq_noise_requirements.png
VCO_freq_noise_requirements.png
Attachment 4: VCO_freq_noise_requirements.pdf
VCO_freq_noise_requirements.pdf
  435   Fri Nov 13 12:54:57 2009 AidanComputingDAQNaming conventions ...

There are now four "sites" in data viewer. [C, C2, C3, C4].

This needs to be fixed. C1 = 40m, C2 = ATF. We need to settle on whether the PSL and TCS labs are the same site and should be identified by subsystem names or are totally different Caltech sites.

"C" by itself is not an acceptable name.

  442   Sun Nov 15 13:50:48 2009 AidanLaserFiberFrequency Counter: BK Precision 1856D

 

I'm not sure when it first arrived but this has been in one of the cupboards for some time. I'm not sure where the network analyzer (PSL lab?) is so I just used this instead. It's got an Adhikari Lab sticker on it. It's also got an RS232 connection on the back which could proved useful.

http://www.bkprecision.com/products/model/1856D/35-ghz-multifunction-counter-frequency-period-totalize.html

  444   Mon Nov 16 11:33:16 2009 AidanComputingDAQUnable to restart framebuilder ...

I can't get the framebuilder restarted because the network topology is now totally different and the old commands now seem obselete. Neither does the 'restart fb0' button in the medm screens work anymore. We need a description of how to restart the DAQ on the wiki. We even have a shiny new page to put the information.

https://vcs.ligo.caltech.edu:448/wiki/wiki?DAQ/Restarting_The_DAQ

 

  445   Mon Nov 16 11:50:27 2009 AidanComputingDAQUnable to restart framebuilder ...

Quote:

I can't get the framebuilder restarted because the network topology is now totally different and the old commands now seem obselete. Neither does the 'restart fb0' button in the medm screens work anymore. We need a description of how to restart the DAQ on the wiki. We even have a shiny new page to put the information.

https://vcs.ligo.caltech.edu:448/wiki/wiki?DAQ/Restarting_The_DAQ

 

The frame builder is running now. It's not at all obvious whether this is as a result of something I did which had a particularly long time constant to initialize or if someone else logged in and got things going.

 

  447   Tue Nov 17 12:17:33 2009 AidanLaserFiberBeam propagation after partial reflector on far side of fiber

Here is the measured beam propagation from the fiber side of the fiber. The ellipticity is coming from the AOM.

Attachment 1: 2009_11_17-beam_propagation_from_PRR.pdf
2009_11_17-beam_propagation_from_PRR.pdf
  449   Tue Nov 17 16:27:17 2009 AidanLaserFiberAdded cylindrical lenses to fiber transmission

Quote:

Here is the measured beam propagation from the fiber side of the fiber. The ellipticity is coming from the AOM.

 I added a couple of cylindrical lenses after the partial retro-reflector on the far side of the fiber. They are placed at the point +/-5mm where the horizontal and vertical beam sizes are the same. The horizontal beam is diverging faster than the vertical. Therefore I used an f=150mm lens in the horizontal and f =250mm lens in the vertical to give the beams approximately the same curvatures at the point where they have the same beam size - hence making the beam roughly circular.

I measured the beam propagation after these lenses. The results are shown below. The beam is now mostly circular.

Attachment 1: 2009_11_17-beam_propagation_from_PRR_post_cylindrical.pdf
2009_11_17-beam_propagation_from_PRR_post_cylindrical.pdf
  450   Tue Nov 17 18:46:36 2009 AidanLaserFiberHeterodyne beat on far side of fiber

I combined the single-pass transmission through the fiber (1x -80MHz and 1x +70MHz) with a pick-off from the NPRO and observed a 10MHz beat on the frequency counter. After demodulation this will be the out-of-loop measurement of phase noise.

  458   Fri Nov 20 10:59:29 2009 AidanLaserFiberOut of loop measurement of stabilized fiber phase noise

Quote:

Good that the loop is closed, but that is an incredibly bad out-of-loop spectrum. Its as if your loop has no low frequency gain.

How about an in-loop noise spectrum on top of the previous plot? What's the loop bandwidth and is it supposed to be 1/f?

 

I think the probably was the loop being slightly unstable before. Previously I had stupidly turned off the in-loop signal on the oscilloscope display so we could just look at the out-of-loop signal and so didn't pick up on this. Trying again with lower gain and in-loop and out-loop time series look much quieter.

Spectra will be forthcoming.

  470   Mon Dec 7 14:01:18 2009 AidanComputingfubarFront-end is down ...

I tried to make a change to the front-end in Simulink and compile it on fb0 - which is supposedly now our front-end machine. For some reason it won't build the .rtl file that is the front-end itself. When I tried to revert to the backed-up model I had the same problem. It doesn't look like anyone has tried to rebuild the front-end since late September and there have been some changes to the network since then.

I'm going to track down Alex and sort this out.

  472   Mon Dec 7 17:18:40 2009 AidanComputingDAQFront-end is back up (Alex's comments)

Quote:

I tried to make a change to the front-end in Simulink and compile it on fb0 - which is supposedly now our front-end machine. For some reason it won't build the .rtl file that is the front-end itself. When I tried to revert to the backed-up model I had the same problem. It doesn't look like anyone has tried to rebuild the front-end since late September and there have been some changes to the network since then.

I'm going to track down Alex and sort this out.

 

Some piece of the code wasn't getting generated properly. /cvs/cds/advLigo/src/fe/atf/atffe.rtl was missing, so the makes did not go as planned. This was fixed by a reboot of fb0 with minor complications:

I tried to reboot fb0 via an ssh, and it wouldn't boot. It kept giving errors (on the monitor on top of the rack) saying something about not being able to find fb1. We power cycled the bottom box (just fb0?), and still nothing. Alex got on the case and fixed it. Apparently the problem is a trash file that gets generated if the RCG model is compiled on a computer that is not running RTLinux. That one was my fault.

Here are his comments:

Email 1:

 

Hi Aidan,

 

Seems to compile fine in ~/advLigo. I did make clean-atf and then make

 

atf. What kind of an error message were you getting?
-alex

My reply:

Hi Alex,
I just tried it again and this is what happens:
$ make install-daq-atf
Installing ... (this all works okay, until)
/bin/cp: cannot stat 'src/fe/atf/atffe.rtl' No such file or directory
Of course when you try 'startatf' the log.txt file in target/c2atf/  
just says:
/cvs/cds/caltech/target/c2atf/atffe.rtl: command not found
Thoughts?
Aidan.
Email 2:

 

 

 


OK, I have it fixed. Have you tried building this on some other computer?
advLigo/src/fe/atf/GNUmakefile got created somehow and it does get
created if you try building on the computer not running the RTLinux
system. I have put some code into the build script to clean this file up
and it the build is working fine now.

-alex

 

 

 

 

 

 

 

 

 

 

  499   Thu Dec 17 19:52:22 2009 AidanComputingDAQframebuilding funniness - "hold please"

Quote:

The framebuilder wouldn't start. Now fixed. This is what I did:

  • I rebooted fb1 via power cycling.
  • Rebooted fb0 the same way
  • did a startatf.
  • did a killatf
  • Couldn't telnet into fb (or fb0) via port 8088 or 8087
  • used the "restart fb0" button in the FE diagnostic

The process seemed to respawn some minutes after this.

 

 

I encountered the same problem this evening. I tried to set another channel to acquire in the frame builder. I uncommented C2:ATF-GENERIC_GEN7_IN_DAQ and set acquire = 1.

When I restarted FB0 from the FE diagnostic screen I just got a message from FB0 saying that the process was spawning too fast and it was going to wait 5 minutes and try again. This repeated many times.

INIT: Id "fb" respawning too fast: disabled for 5 minutes.

Right now it is still trying to start fb again.

  500   Thu Dec 17 20:40:35 2009 AidanComputingDAQRestarting DAQ - something else to avoid

I had some issues restarting the FE. It turned out there was a typo in C2ATF.ini so one of the data rates was 2^N. I fixed that and the FE started nicely again. I even locked up the PMC again.

 

  533   Sat Jan 16 23:16:45 2010 AidanComputingDAQRestarted front end ... and then it broke

I switched some channels to acquire for the fiber stuff. That is, after saving the current version of C2ATF.ini as C2ATF.ini.old100116, I edited the file to set the follwing channels to acquire: C2:ATF-GENERIC_GENx_IN1_DAQ and C2:ATF-GENERIC_GENx_OUT_DAQ for x = 1:7. Unfortunately, I meant to reboot the framebuilder and I accidentally restarted the whole front-end. I rebooted fb0 and did a startatf but now the front-end is not responding - at least, all of the EPICS displays corresponding to the DAQ inputs are showing 0 rather than fluctuating values.

It's been a really long day and I'm too tired to try fixing this right now.

  535   Mon Jan 18 17:53:13 2010 AidanComputingDAQFront end restarted - "burtrestore = 1" was the trick

Got the front-end going again. It turned out to be straight forward. I tried the following:

  1. Restored C2ATF.ini from the backed up version
  2. Logged into fb0 and rebooted: [controls@fb0 ~]$ sudo reboot
  3. Once fb0 was back up I logged back in as controls and started the front end: [controls@fb0 ~]$ startatf
  4. The EPICS channel displays went from white to showing all zeros. 
  5. On the C2ATF_GDS_TP.adl screen (the FE Diagnostic screen) I switched burtrestore from 0 to 1 and this brought all the channels back up to full operation
  536   Tue Jan 19 09:17:09 2010 AidanComputingDAQNew fiber channels in frame

I added the following channels to C2ATF.ini

 

# The follwoing channels were added by Aidan Brooks on 18th Jan 2010
# They are just C2:ATF-GENERIC_GENx_IN1_DAQ and GENx_OUT_DAQ for x =
# 1:7
# GEN1_IN1 - BNC CH9 on Anti-Alias Chassis
[C2:ATF-FIBER_PD_OUTLOOP_DC_IN1]
datatype=4
chnnum=10241
acquire=1
datarate=8192
[C2:ATF-FIBER_PD_OUTLOOP_DC_OUT]
chnnum=10243
acquire=1
datarate=8192
datatype=4
# GEN2_IN1 - BNC CH10 on Anti-Alias Chassis
[C2:ATF-FIBER_PD_INLOOP_DC_IN1]
datarate=8192
acquire=1
datatype=4
chnnum=10244
[C2:ATF-FIBER_PD_INLOOP_DC_OUT]
datatype=4
datarate=8192
chnnum=10246
acquire=1
# GEN3_IN1 - BNC CH11 on Anti-Alias Chassis
[C2:ATF-FIBER_REF_BEAM_PWR_IN1]
chnnum=10247
datatype=4
acquire=1
datarate=8192
[C2:ATF-FIBER_REF_BEAM_PWR_OUT]
datatype=4
datarate=8192
acquire=1
chnnum=10249
# GEN4_IN1 - BNC CH12 on Anti-Alias Chassis
[C2:ATF-FIBER_TRANS_BEAM_PWR_IN1]
chnnum=10250
datatype=4
datarate=8192
acquire=1
[C2:ATF-FIBER_TRANS_BEAM_PWR_OUT]
datarate=8192
datatype=4
acquire=1
chnnum=10252
# GEN5_IN1 - BNC CH13 on Anti-Alias Chassis
[C2:ATF-FIBER_PD_OUTLOOP_AC_IN1]
datarate=8192
acquire=1
datatype=4
chnnum=10253
[C2:ATF-FIBER_PD_OUTLOOP_AC_OUT]
datatype=4
datarate=8192
acquire=1
chnnum=10255
# GEN6_IN1 - BNC CH14 on Anti-Alias Chassis
[C2:ATF-FIBER_PD_INLOOP_AC_IN1]
datatype=4
chnnum=10256
acquire=1
datarate=8192
[C2:ATF-FIBER_PD_INLOOP_AC_OUT]
datarate=8192
datatype=4
chnnum=10258
acquire=1
# GEN7_IN1 - BNC CH15 on Anti-Alias Chassis
[C2:ATF-FIBER_DBL_TRANS_PWR_IN1]
chnnum=10259
datarate=8192
datatype=4
acquire=1
[C2:ATF-FIBER_DBL_TRANS_PWR_OUT]
chnnum=10261
datarate=8192
datatype=4
acquire=1

  537   Tue Jan 19 09:18:47 2010 AidanLaserFiberDouble passed stabilization - mark II

I closed the loop again on the fiber stabilziation. The attached times series shows the out of loop noise being surpressed substantially.

The channels are as follows:

 

Legend for attached plot.
Channel name Description
C2:ATF-FIBER_PD_INLOOP_DC_IN1 The DC output from the in-loop RF PD [New Focus 1811] (in counts)
C2:ATF-FIBER_PD_INLOOP_AC_IN1 The AC output from the in-loop RF PD (in counts)
C2:ATF-FIBER_PD_OUTLOOP_DC_IN1 DC output from out-loop RF PD [New Focus 1811] (in counts)
C2:ATF-FIBER_PD_OUTLOOP_AC_OUT AC output from out-loop PD (counts) - whitened (analogue) and anti-whitened (digital) this is the most important channel
C2:ATF-FIBER_REF_BEAM_PWR_OUT DC value of the reference beam pick-off power (in uW)
C2:ATF-FIBER_TRANS_BEAM_PWR_OUT DC value of the fiber transmission beam pick-off power (in uW)
C2:ATF-FIBER_DBL_TRANS_PWR_OUT DC value of the double-passed fiber transmission pick-off power (in uW) *contains a significant amount (~30-50%) of back-scatter in addition to double-passed transmission.


Loop details to follow ...

 

Attachment 1: 2010-01-19_loop_closed_times_series_good.pdf
2010-01-19_loop_closed_times_series_good.pdf
ELOG V3.1.3-