The quantities we want to measure as a function of the temperature:

- Temperature: 2.2k thermister resistance / 100ohm platinum RTD

- QE (Illuminating output / Dark output / Reference voltage / Reference dark output)

- Dark current (vs V_bias) -> Manual measurement or use a source meter

- Dark noise (PSD) 100kHz, 12.8k, 1.6kHz, 100Hz

I borrowed KEITHLEY 2450 source meter from Rich. The unit comes with special coaxial cables and banana clips. Most of the peripherals are evacuated in the OMC lab.

The dark current of A2P2, A2P3, A2P6 were measure with different temperatures (300K, 270K, 254K). The plot combined with the previous measurement ELOG QIL 2425.

== How to use the source meter ==

- Two-wire mode: Connect the wires to the diode

- Over voltage protection: [MENU] button -> SOURCE / SETTINGS->Over Voltage Protectiuon 2V

- Sweep setting: [MENU] button -> SOURCE / SWEEP -> e.g. Start -750mV, Stop +500mV, Step 10mV, Source Limit 1mA -> Select Generate

- Graph View: [MENU] button -> VIEWS / GRAPH

- Start measurement: Note: The response of [TRIGGER] button is not good. You need to push hard
  This starts the sweep, or a menu shows up if your push is too long -> Select "Initiate ..."

- Data Saving: [MENU] button -> MEASURE / READING BUFFERS -> Save to a USB stick

[Raymond, Aidan, Chris, Koji]

P6 element (500um)^2

- We looked at the current amp (FEMTO) output. The amplifier saturated at the gain of 10^3 V/A. Looking at the output with a scope, we found that there is a huge 1.2MHz oscillation. Initially, we thought it is the amplifier oscillation. However, this oscillation is independent of the amplifier bandwidth when we tried the our-own made transimpedance amp.

- Shorting the cryostat chamber to the optical table made the 1.2MHz significantly reduced. Also, connecting the shield of the TEC/Laser controller made the oscillation almost invisible. This improvement allowed us to increase the amp-gain up to 10^7.

- Then the dominant RMS was 60Hz line. This was reduced by more grounding of the cable shields. The output was still dominated by the 60Hz line, but the gain could be increased to 10^8. This was sufficient for us to proceed to the careful measurements.

----

- The dark current was measured by the source meter, while the photocurrent (together with the dark current) was measured under the illumination of the ~1mW light on the PD.

- Attachment 1 shows the dependence of the dark current against the swept bias voltage. We had ~mA dark current at the room temp. So, this is ~10^5 improvement.

- Attachment 2 shows the dependence of the apparent QE against the swept bias voltage. The dark current was subtracted from the total current, to estimate the contribution of the photocurrent in the measurement.

- Attachment 3 shows the dark noise measurement at the reverse bias of ~0.6V. Up to 1kHz, the noise level was below the equivalent shotnoise level of 1mA photocurrent.

---------

All the data and python notebook in the attached zip file.

The QE measurements from the first couple of photodiodes are attached below.

• plot_JPL_diode_results.m - A2P6 analysis
• plot_JPL_A2P2_diode_results.m - A2P2 analysis

QE = [I_photocurrent]/[P_PD] * h *nu/e

P_PD = Power incident on photodetector = 0.966*power_incident on cryo window

Power incident on cryo window = F(voltage on reference PD)

Item lending as per Ian's request: Particle Counter from OMC Lab to QIL

The current particle class of the room was measured to be 800.

The particle counter went back to the OMC lab on Aug 10, 2020.

Still trying to figure out how to set up the particle counter remotely. The current particle count is 576.

Note: the particle count is the number of particles detected over 0.3um size.

West Bridge flooding Apr 6th due to rain in the night

Looks like the first responder was Calum. The attached photos were sent from him.

We can use Thorlabs SAF1450S2 gain chip to generate 1418 nm light using an ECDL design similar to the one described in Kapasi et al. Optics Express Vol. 28, Issue 3, pp. 3280-3288 (2020) (ANU 2um ECDL design).

PZT Transfer function

I have contacted Disha and Johannes to get the actual measured data for the PZT transfer function of this ECDL design. Fig.5b in their paper plots the transfer function of the PZT. Since, in ECDL PZT directly changes the cavity length, it has a more powerful actuation strength (2 orders of magnitude more) with actuation of 560 MHz.V upto 1 kHz. It however had a very low pole at 1 kHz and two mechanical resonance-antiresonance pairs near 1 kHz and 2 kHz. I modeled a transfer function by eye using Fig.5b of the paper. Page 1 in the attached pdf shows this modelled transfer function.

AUX PDH Loop

Next, we need to change the PDH loop for the auxiliary laser lock with the 40m cavity since the PZT has changed. I modelled one from scratch. This simple analog loop's performance is shown in orange in pages 2-5. This loop seemed stable from all the metrics I know, viz: phase margin of about 55 degrees (Page 2), no strong peak in close loop transfer function (page 3), and no remanant oscillations in time domain response (page 4).

I also modeled a similar loop but with digital compensation of the resonance-antiresonance features. This loop is plotted in green on pages 2-5. Both these loops have 300 kHz of bandwidth just by using PZT. I beleive this could be increased but I have not taken into account any saturation of PZT.

ECDL Frequency Noise

From Fig.4. of the paper gives a frequency noise estimate for free running ECDL. They mentioned that a roll-off below 10 Hz was due to their thermal feedback to remain in linear range of their frequency noise emasruement method. I modeled the noise of ECDL hence by

where the flicker noise contribution is similar to NPRO noise but ECDL has a white noise of 15 Hz/rtHz due to natural linewidth of spontaneous emission or Schawlow-Townes linewidth (with several broadening factors). I think this is an inherent limitation of ECDLs.

Page 5 shows both unsuppressed and suppressed frequency noise estimate for ECDL with the loops mentioned above and current values of NPRO noise are also plotted for comparison.

Code

Following up on the last post, here I presented a near back of the envelope calculation of how different choices of AUX cavity finesse and laser source for mariner would affect the prospects of calibration scheme.

Laser sources considered:

As mentioned in the last elog post, here I considered using an NPRO seeded auxiliary laser source (converted to 1418nm by whatever method), ECDL based on ANU design with a modified PDH loop and same ECDl with a digital compensation of PZT resonances. I have taken the residual frequnecy noise of these lasers as the dominant noise source in the calibration scheme. Craig and Gautam in their proposal for SoCal wanted the AUX laser to be locked to the arm cavity in a PDH shot noise limited way. That would be necessary for 4km interferometers and would be easier to achieve there with higher laser powers and higher cavity finesse.

Finesse of 40m Arm Cavity for 1418nm:

Here I considered three cases. First assumes about 3% transmittance of 1418nm in ITM and ETM HR coatings for mariner. This gives a finesse of about 100 and a cavity pole of 18.9 kHz. I believe this is the existing case at 40m. Next we consider transmittance of 0.5% and 0.05% (500 ppm) of 1418nm in ITM and ETM HR coatings for mairner. These cases give finesse of 625 and 6.28k respectively with cavity poles at 3 kHz and 299 Hz respectively.

Page 1: Consideres the case of finesse of 100. The green dashed line shows the amount of drive strength (in m) required at different frequencies if we use ECDL with PZT resonance compensation, to get an SNR of 1000 in 100s of integration time.

Page 2: Same as above but for Finesse of 625.

Page 3: Same as bove but for Finesse of 6280.

Page 4: Comparison of different finesse cases for the ECDL with PZT compensation option. Dashed curves represent requried drive strength (in m) for different cases.

Page 5: Same as above but for NPRO seeded auxiliary laser.

Note: For the NPRO seeded auxiliary laser, we have assumed that the noise of conversion to 1418 nm is similar to noise due to SHG process which is not dominant. There would be an effect of multiplying with a factor ranging form 1-1.5 due to frequency conversion but I have ignored it here for simiplicity. Also, NPRO case is limited in bandwidth due to PZT resonances. We might be able to get away with them using digital compensation like the case study for ECDL. But I haven't attempted that here as we do not know our NPRO PZT's resonance features yet.

I don't know why it would be easier to have higher finesse with longer arms. Something about beam size???

The NPRO PZT TF's are all in the 40m elog - there are many measurements of TF made over the past 10 years. Its like Raiders of the Lost Ark - you have to believe its there while searching.

If we use ECDL for auxiliary frequency in 40m and hope to stabilize it up to 1 MHz with digital compensation of PZT, it is important to take into account any phase effect of the nearby FSR at 3.97 MHz. This should ideally be included in the Input Mode Cleaner loop considerations as well. These effects would be more prominent in longer cavities like aLIGO and LISA where FSR is very low and should we attempt to stabilize a laser lock beyond cavity's FSR.

I did a no assumptions calculation for getting a general transfer function fo PDH error signal in units of [W/Hz] assuming 1 W of incident power. This calculation would soon be uploaded here. I'll put down here primary results.

For incident field on a Fabry-Perot cavity (with fsr of ), reflected electric field transfer function (unitless) is given by:

Then, PDH error signal for a modulation frequency of at a modulation index of , in units of [W/Hz] (i.e. error signal power per Hz of error in laser frequency from cavity resonance) is given by:

after demodulation and low pass filtering. Note this transfer function is a complex quantity as it carries phase information of the transfer function too. The real signal is obtained by multiplying this signal at with and taking the real value of the product.

Having done this, we can see how in the real PDH error signal, there is a low pass at cavity pole, given by and a notch every fsr. The notch creates a zig-zag in the phase of the tranfer function and has a HWHM same as cavity pole. After this point, I just fitted a ZPK model to the transfer function to obtain a empirically derived model for PDH error signal transfer function. Apart from the cavity pole, this model needs to have resonance and antiresonance features present at each FSR with resonance having a linewidth of cavity pole while anti-resonance having a linewdth of . Here's how the ZPK model would look like:

I've attached my notebook where I did the fitting analysis and the overlap plot of real PDH error signal TF and the modelled approximation.

I have a preliminary calculation to post here. This does not include noise sources from cavity fluctuations and main frequency noise. But it gives some idea about shot noise and frequency noise of AUX laser conttribution to the noise in calibration.

What's included?

• I have put in measured PZT transfer function of ECDL at ANU upto 25 kHz. Above this point, they did not measure it so I couldn't make it artificially dirty. I just assumed 1/f roll off above (which is definitely incomplete picture).
• I have included phase effects of cavity FSR in the loop by adding the resonance features as mentioned here.
• I have added attempted resonance compensation for 1kHz and 2kHz features in the PZT after fitting the data with poles and zeros and iverting them.
• 10mW of incident AUX light is assumed on the arm cavity.
• Total 10 mW of combined power at 709 nm is assumed to fall on the beatnote. So AUX light would be frequency doubled for this beatnote.

What's left?

• Need to add seismic noise and other measured excess noise that come from the cavity motion.
• Need to add laser frequency noise of main laser, however, it must be small since it is locked to mode cleaner.
• Need to add digital delay of Red Pitaya or whatever filter would be used for PZT resonance compensation.
• Need to model PZT transfer function of ECDL above kHz properly. ANU replied that they can't measure it for higher frequencies due to lack of time.
• Need to do time domain stability analysis. This I haven't been able to do as I have just been using python-controls package as black box to compute impulse and step responses of state space systems. When simply adding easured transfer function data, I couldn't create the state stpace representation for the system. I tried to fit multiple resonances above 2 kHz but couldn't really capture the magnitude of the response well. Maybe I can just assume higher harmonics of the 1kHz and 2kHz resonances?

Attachments:

• Page 1 is the measured PZT transfer function fo 1900 nm ECDL from ANU along with the modeled 1/f roll off after 25 kHz.
• Page 2 is the open loop transfer function of the AUX PDH loop for the three different finesse cases studies. Note that the blue curve is hidden beneate the other two curves. Before objections came, I know this is unreal and incomplete, but I have to start somewhere.
• Page 3 is the calibration noise budget with different colors showing the three finesse cases.
  • This is also incomplete but we can takeaway what the shot noise contribution would look like and initiate a dialogue about the integration time chosen (which is 100s here), the SNR aim chosen (1000 here) and what drive strength would be good enough.
  • From notes of Craig and Gautum, I think we can drive the mirrors at 0.1pm ampltude in the calibration band. From my first and only calibration measurement in 40m, I could drive upto a pm without loosing lock in the cavities.
  • But that was a simple single arm lock and full ifo lock must be more sensitive. So is 0.1 pm drive strength good enough or do we want to aim lower?

An issue was raised with last calculation about the fact that our sensing of PDH signal isn't ideal and in the real world there is scattering, clipping extra adding excess noise in the PDH loop. This noise primarily comes by the intensity noise imparted on promptly reflected light from the cavity via various shaking optics etc on the table before it goes to the PDH reflection RF photodiode.

This noise's coupling to the PDH loop is identical to how shot noise of light couples into the PDH loop i.e.:

• Intensity noise of light is converted into voltage oise by the PDH photo diode.
• This is compared against the cavity finesse amplified real PDH error siganl at this stage.
• Therefore, in frequency noise, the affect of this intensity noise is smaller for higher finesse cavities since cavity finesse only amplifies the PDH signal anf not the scattering noise.

Excess noise estimate

• I used this measurement taken in 40m with Koji to estimate this noise.
• This measurement contained a beatnote between IR coupled AUX light and the main laser IR pick-off when X-arm is locked to the main laser and AUX laser is locked to X-arm.
• So this noise measurement is an upper bound on the total noise in AUX laser frequency when it is locked to the X-arm.
• I compared this against the noise budget model for AUX PDH loop I have which uses the same control loop as the uPDH box used here.
• I found a bulge of excess noise below 100 Hz and it seemed to go done as 1/f^2 there. I was reminded by a chat I had with Rana and another professor sometime last year when Rana mentioned scattering noise showing up as "Scatter shelf" looking something like this.
• So I modeled excess noise as the difference between the noise budget and the measured noise with it extending after 100 Hz with the same roll off as in 10-100 Hz.

Calibration noise budget

• I took the excess noise measured, converted it to W/rtHz by using current AUX PDH discriminant and photodiode gain, and normalized by the power (9.6mW) to get this noise in RIN/rtHz.
• Then I assumed that the same RIN would be imparted in the Mariner AUX loop and calculated excess intensity noise at the PDH loop by multiplying the above number with assumed 10mW of incident power to get it in W/rtHz.
• From here, I fed it to the same input as I feed the shot noise in the loop and calculated the effect in the overall noise budget.
• For high bandwidth and gain PDH loops required for calibration, this kind of noise would dominate up to a kHz before getting taken over by the residual laser frequency noise.
• I have again plotted cases for three choices of finesse/mirror transmission. If we used 99.95% reflectivity (1000 ppm transmission, finesse of 3140) we would be fine for most calibration lines except the one around 40 Hz. (assuming drive strength of 1e-13 m everywhere).
• Otherwise, if possible, we should go for higher finesse. Case (c) plotted here (Page 3), shows that for 99.995% reflectivity (100 ppm transmission, finesse of 31415), we will be fine in all over the range with cavity pole dropping to 63 Hz. This would be really nice of course, if it is possible.
• So we recommend HR coatings for 1418 nm in the Mariner to be 99.995% reflective (giving total power transmission of 100 ppm).

I moved the brand new TED200C on the workbench to Crackle for 2um ECDL (permanently)
The TED200C temp controller used in the 2um PD test setup will stay there (permanently)

http://nodus.ligo.caltech.edu:8080/SUS_Lab/1851

FEMTO DLPCA200 low noise preamp (brand new)

Keithley Source Meter 2450 (brand new) => Returned 11/23/2020

were brought to the OMC lab for temporary use.

I was thinking about getting this new current pre-amp from NF:

http://www.nfcorp.co.jp/english/pro/mi/loc/pre/ca5351/index.html

It seems to have a good noise performance and has a built in low pass filter and also a remote interface.

The FEMTO seems less fancy, but their noise performance is actually 2-3x better.

Is the reverse bias programmable? FEMTO has a bias trimmer on it. It's useful in the usual application, but for automation, the configuration of the input becomes cumbersome.

doesn't seem so, but they sell this one:

http://www.nfcorp.co.jp/english/pro/mi/loc/m_lp/p_lp/lp_6016_01/index.html

which has a USB interface and pretty good voltage noise spectrum

I'm attaching my rough first draft of the QIL photodiode testing schematic. Please provide comments for fixes/improvement!

Posting link to PD testing google doc here: 

https://docs.google.com/document/d/1RrtX5nqNeEOazNvT2rPxHnSqumgvYfaBrYTMMGZ-ecc/edit

Attached:

- Updated schematic of the current PD testing setup, including noise levels for current electronics

- Table of desired measurements for new setup, with expected signal levels, accuracy, and readout values

Looks very clear, thanks. I guess the next thing to do is

1. ask if this will work for all the various PDs we want to test,
2. is it good enough for all our requirements, and then we
3. draw a new diagram for the new setup, incorporating what to keep and what circuit to make ourselves

Note that the back panel connectors are Triax, not the usual Coax.

I tried pumping down the JPL PD chamber to test the new PD at cryo temperatures. Unfortunately, the chamber can;t get past about 6E-3 Torr with the pump on. As soon as I turned off the pump the pressure rose to around 2 Torr over 20 minutes or so.

I extricated the chamber from the pedestals, flipped it and removed the bottom plate. I cleaned the O-ring with isopropanol and wiped down the mating surface on the chamber (also with iso). I replaced the plate and tightened the screws. Then I returned the chamber to the table and reconnected it to the vacuum system. I tried pumping down once again but I saw pretty much exactly the same situation as before (pressure bottoming out around 6E-3 Torr and then rising quickly again when the pump was turned off).

I guess it's possible that the O-ring is damaged - although I couldn't see anything obivous. We didn't mess around with the viewport (when we replaced the diode a few weeks ago) so I'm hoping there is no issue there.

I tried Krytox around the O-ring and also tightening the screws around the valve. The leaking persists at roughly the same rate.

[Stephen, Aidan, Wednesday 09 June]

Summary and Plan:

• Poor sealing at the output valve (The Gap) needs to be resolved.
• Planning to install #4-40 helicoils today (chamber will remain sealed, will need to remove output valve and cover output orifice, then transport the chamber to the WB EE shop for redrilling of holes.)
• Meeting with Nina and Aidan this afternoon to iterate one more time.

Troubleshooting steps taken:

1. Aidan took us through the full sequence of pump down and disassembly to bring me up to speed.
2. We opened the lid and inspected the old o-ring.
   1. Signs of plastic deformation and of small flecks of particulate near sealing surface - good idea to change.
3. We found new o-rings in a box from the Cryo lab, and one of these was swapped in after a good wipedown with IPA.
4. Upon pumpdown, Aidan compared behavior and found no meaningful change to rate of pumpdown or stable pressure in e-4 torr range after 10+ minutes.
   1. By valving off [chamber + gauge] from pump line, it was clear that there was a leak in that volume, as within seconds the pressure rose from e-4 torr to e-2 torr, and stabilized at _(need to confirm - e0?)_ torr over ~10 minutes.
5. Attempted to squirt IPA along o-ring seals, but there was not good access to the sealing surfaces, so this was a null test
6. Looked closer at all of the chamber features, and noticed The Gap between chamber wall and chamber output valve, pictured in [Attachment 1]. Not good! But promising as a leak source.
   1. Three of the four screws were found to be loose due to apparent thread damage.
7. IPA was squirted into The Gap at stable pressure of e-4 torr, but no change in pressure was noticed.
8. Longer #4-40 screw reinstallation was attempted, and I could feel a small amount of pull at the very tip of the screw, but tightening the screws led to that small pull to fail as well - need to rework.
9. The Clamp was installed and The Gap was closed [Attachment 2].
   1. When isolated from the vacuum pump, the chamber pressure progressed more slowly. Within seconds, we were at e-3 torr, and over 10 minutes the pressure stabled at e-2 torr, about 30x lower pressure per Aidan's records.

We hit another dead-end with leak hunting the IR labs dewer (we replaced screws and helicoil on the valve connection but there is still a big leak). We cleaned the flange and O-ring with isopropanal and replaced the threads with helicoil but still get the same sort of leak where we only hit 1E-2 Torr after 5 minutes of pumping and stablize around 1E-3.

After turning off the pumping station, the pressure rose quickly to 1Torr (in roughly 10 minutes or so).

I wrote a python script to extract temperature data from the CTC100 via ethernet, for monitoring cooldown/warmup of Megastat. This is intended to replace USB data extraction, which requires the user to manually insert/remove the stick and plug into a computer.

The script queries the CTC100 every ~60 seconds for the latest temperature values (the frequency can be supplied as a parameter, but default is 60s). The script writes line-by-line to a .txt file and also plot the outputted data once collection is terminated.

Here is a gitlab link to the script: https://git.ligo.org/voyager/mariner40/-/blob/master/CryoEngineering/ctc100_controller.py. It is also found on the QIL workstation at /home/controls/CTC100/ctc100_controller.py. To run from the workstation, open terminal to /home/controls (home). Then:

 Here, 'tutorial' stands in for the desired filename for the outputted data. The script will start pulling data and will print each line to the terminal. It will continue printing and logging the temperature values until the user hits Ctrl+C in the terminal. This will terminate the script and output the final data file. The file is saved as a .txt file in /home/controls/CTC100/data.

*Takeaway*: The current 1D cooling model is getting closer to matching our observed cooling trends, mainly in the lower temperature limit. The predicted time constant is still much smaller than we are seeing in reality (by about a factor of 3), but this can potentially be improved by revising specific heat values and/or dimensional estimates for chamber components.

The model uses the known cooling power of the cold head [attachment 2] and considers radiative heat from the outer shield, baseplate (bottom lid), and mylar wrapping around braid. I increased the complexity of the script by solving a system of ODEs (for braid and coldplate temperature) simultaneously instead of assuming the temperatures are equal at all times, and solving only 1 ODE. This resulted in the model's lower temperature limit prediction matching our observed data, at ~66 K. 

The model still predicts a much smaller time constant than we are seeing. This is affected by specific heat values for Cu and Al, along with dimensional estimates of the coldplate and braid (AKA how much mass is being cooled). It is possible that these values are being underestimated in the model, which would lead to the smaller time constant. Currently the model uses constant values for the specific heat of Cu and Al (room temperature). But since specific heat increases with temperature, accounting for temperature dependence would lower the specific heat values and shift the model in the opposite direction (towards an even smaller time constant). Therefore I suspect the model is underestimating the mass of the coldplate, though I am unsure if this would completely correct the discrepancy. 

If the term (specific_heat * density * volume) of the coldplate (Al) is increased by a factor of 4, the model resembles the data well [attachment 3]. 

[Stephen Koji Radhika]

Stephen and Radhika worked on the cooling down and warming up of the cryostat with the cold head RTD attached using a spring-loaded screw. No other configuration changes compared to QIL/2599. Here are the temperature log plots. Photos of spring clamped RTD are outstanding, but the clamp is the same as the workpiece pictured in QIL/2599/Attachment 12.

[Stephen / Koji]

Bonding work for the prep of the preliminary suspension test

- 1" sq mirror-ish polished SUS piece was bonded to a face of the silicon mass. We chose the location right next to a line on the barrel. (Attachment 1)

- The mass was flipped with two more same thickness pieces used for the spacers to keep the mass horizontal.

- A pair of an OSEM and dumbbell-magnet was brought from the 40m (courtesy by Yehonathan). The magnet was glued on the mass at the opposite position of the attached mirror because the optical ports are going to be arranged to share an axis. A piece of cryo varnish was also painted with a piece of cigarette paper at the center of the mass so that we can attach an RTD. (Attachment 2)

Next Things To Do (Attachment 3)

• Vent the chamber
• We will move an optical port to the opposite position of the other port.
• A DB9 feedthru is going to be installed.
   
• Suspension
  • Move the sus frame in the chamber
  • Suspend the mass
• Sensor arrangement
  • Set up the oplev
  • Hold the OSEM at the height of the magnet
  • Set up a camera to observe the magnet-OSEM clearance
  • We improvise the DB crimping sockets so that we can electrically connect the OSEM (optional)
• Pump down / cool down the chamber
  • The main target of the cooling is to check the cooling capability of the test mass mainly with radiative cooling.
  • An optional target is to observe the misalignment as a function of the temperature -
    • -> Oplev signals are to be connected to CDS / check if CDS is logging the data
  • Check if the OSEM/magnets survive the thermal cycle
  • If possible we can try to actuate the OSEM / check the LED/PD function at the cryo temp
     

[Stephen Koji]

We started cooling down of the test mass.

Venting

- Stephen vented the chamber at 2PM. An optical port was moved to see the OSEM from the back.

OSEM wiring

- Brought DSub crimp sockets from the 40m. We picked up 3x 1m LakeShore WCT-RB-34-50 (twisted silver-plated copper, 34 AWG with Teflon insulation). The ends of the wires were dangled so that crimping is possible. A single wire resistance was measured to be ~1Ohm at room temp. (Attachment 1)

- OSEM pin out / backside view (cable going down) (Attachment 2)

|   o   o   o |
| o   o   o   |                 Wire
  ^ ^ ^ ^ ^ ^---PD K        ---- R3
  | | | | |-----PD A        ---- B3
  | | | |-------LED A       ---- B2
  | | |---------LED K       ---- R2
  | |-----------Coil End    ---- B1
  |-------------Coil Start  ---- R1

Twisted Pair 1: (R1&B1) with 1 knot  at the feedthru side
Twisted Pair 2: (R2&B2) with 1 knot  at the feedthru side
Twisted Pair 3: (R3&B3) with 1 knot  at the feedthru side

Dsub feedthru in-air pinout (Mating side)

    1  2  3  4  5
\ o  o  o  o  o /
 \ o  o  o  o  /
   6  7  8  9

Pin1 - Coil Start
Pin6 - Coil End
Pin2 - LED K
Pin7 - LED A
Pin3 - PD A
Pin8 - PD K

Pin1-6 R=16Ohm
Pin2-7 Diode V (with Fluke) 1.18V (Pin2 black probe / Pin7 red probe)
Pin3-8 Diode V (with Fluke) 0.7V (Pin3 red probe / Pin8 black probe)

- OSEM pin out / backside view (cable going down)

Suspension installation (Attachment 3)

- The sus frame was moved into the chamber

- We measured the test mass dimension before installation: L 3.977" D 4.054"

- The attached mirror size is 1"x1" made of SUS #8 (?)

- The mass was suspended. The height / rotation of the mass was adjusted so that the reflecting mirror is visible from the oplev window and also the OSEM magnet is visible from the OSEM window.

- The OSEM was placed on an improvised holder. (Attachment 4)

Oplev installation

- ...Just the usual oplev installation. Adjusted the alignment and the return beam hits right next to the laser aperture. This beam was picked off by a mirror and steered into a QPD. (Attachments 5/6)

- The lever arm length is ~38" (960mm) -- 9" internal / 29" external
- The oplev signal is shaking so much and occupying ~50% of the full scale. Added a lens with f=250 to make the beam bigger, but the improvement was limited.

Pumping down

- Started ~8:30PM?

DAQ setup

- Wired 3 BNC cables from the table to the DAQ rack. CHX/Y/S are connected to ADC16/1718ch.

- The real-time processes seemed dead. Looked at [QIL ELOG 2546] to bring them up. TIM/DAQ error remains, but the data stream seems alive now. Leave it as it is.

Cooling

- Temp Logging started. Filename: temp_log_cool_down_20210716_2255.txt

- Cryocooler turned on. ~10:55PM

- Confirmed the cold head temp was going down. The cold head temp is 75K at 0:30AM

OSEM photo

- An example photo was taken from the rear window. The attempt with 40m's Canon failed. Attachment 7 was taken with KA's personal compact camera with a smartphone LED torch. The gap between magnet and OSEM is highly dependent on the view axis. So this is just a reference for now.

Temperature log for the first 2 hours (Attachment 1)

I wonder why the temperatures displayed on CTC100 and the ones logged are different...?

Uh oh, review of the cooldown plot from the previous cooldown (QIL/2603) shows workpiece temperature of ~92 K at conclusion, while a temperature of 65K was observed in the CTC100 readout (Attachment). The logging of the warmup is consistent with the CTC100 image, as the logging started a few minutes after the warmup was started, and the warmup "5 minutes after starting" temperature of ~ 71 K is a practical temperature.

Seems to be something weird going on here, we will need to have Radhika take a look on her return (and continue taking photos of the CTC100 whenever we stop by).

Temp Log on Jul 19 2021 17:20

I wonder what is the heat transfer mode for the test mass right now. Radiative? or Conductive through the wires?

A naive cooling model was applied to the cooling curve.
A wild guess:
- The table temp is the same as the test piece temp as measured on 2021/7/9
- The inner shield temp is well represented by the table temp
- The specific heat of Si is almost constant (0.71 [J/(g K)] between 300K~200K

Radiative cooling:
The curve was hand-fitted by changing the emissivity of the inner shield and the silicon mass. I ended up having the same values for these to be 0.15.
Surprisingly well fitted!

Conductive cooling:
The conductive cooling through the wire does not fit the cooling curve, although the quantitative evaluation of the wire conductivity needs to be checked carefully.

Appendix:
Stephen shared attachments 2 and 3, which contain insights on the wire used to hang the Si mass. .017" diameter Music Wire from California Fine Wire, 2004 vintage, borrowed from Downs High Bay.

Updated the model the latest log data with cooling prediction

• The radiative cooling is expected to be the dominant cooling mode.
• It will take ~3 more days to reach 123K. We don't need to wait for it.
• For more informative temp data, we need the temperature of the inner shield and the table.

• We know the cold head temp from the measurement. For the prediction, the constant cold head temp of 65K was assumed.
• The table temp was estimated using conductive cooling model + linear empirical dependence of the conductivity on the temp
• The constant specific heat of the silicon mass (0.71 J/K/g) was assumed. This may need to be updated.
• The radiative cooling is given from Stefan–Boltzmann law with the emissivity of 0.15 for both the shield and the mass.
   
• The conductive cooling of the test mass was estimated using: Wire diameter 0.017" (=0.43mm), 4 wires, length of ~10cm (guess), no thermal resistance at the clamps (-> upper limit of the conductive cooling)

Radiative cooling already gives us a good agreement with the measured temp evolution for the test mass. The conductive cooling is not significant and does not change the prediction.

Updated the plot with the new data (2021/7/21 12:30PM)

See https://nodus.ligo.caltech.edu:8081/40m/16250

[Stephen and Koji for discussion / Koji for the execution]

1. Temperature Trend

See [QIL ELOG 2611] for the updated temp log and the cooling model.

Considerations for the next cycle:
-> How can we accelerate the cooling? It seems that the table cooling is conduction limited. Improve the cold head connection.
-> We want to move the RDTs
-> How can we improve radiative cooling?

2. Oplev Trend (Attachment 1)

Sum: The beam has been always on the QPD (good). See also Attachment 2

X&Y: In the first few hours the beam drifted in -X and then +X while Y had slow continuous drift in +Y. ~11hours later sudden drift in -Y and totally saturated. Also -X saturation observed @~16hrs. Again +Y drift was seen @~25hrs. The totally saturated in -X and +Y.
They may be related to the drift of various components with various cooling time scale.

Visual check: ~2mm shift in X&Y is visually observed. Attachment 2

-> How can we quantify the drift? What information do we want to extract?

3. OSEM and the magnet

The magnet is intact. And the suspension seemed still free after cooling (Attachment 3)
Significant misalignment was not visible. No visible damage by cooling was found. The coil is alive and the PD/LED are also intact. Fluke showed that they are still diodes, but their function was not checked.

The coil resistance changed from 16Ohm -> 4.2Ohm. For the 16Ohm, 2 Ohm was from the wire. Let's assume we still have 2Ohm overhead -> The coil R changed from 14->2.2. This corresponds to the coil temperature of the order of ~100K. This is not so crazy.

Some actuation current was applied to the magnet. For this test, the oplev was realigned.
First, some ~300mA current pulses were applied to the coil. The ringdown of the yaw mode was visible. Then the DC current of 100mA was applied. This didn't make visible change on the spot position but the data showed that there was a DC shift.

-> We prefer to have a softer suspension for the next test.

4. CTC100 logging

During the cooling we kept having inaccurate data logged compared with the displayed data on the screen of CTC100.
As soon as the cooling logging was stopped, telneting to CTC100 was available. So, I telnetted to the device and sent the data transfer command ("getOutput"). Surprisingly, the returned values agreed with the displayed values.
So my hypothesis is that somehow the data strings are buffered somewhere and gradually the returned values get delayed. From the behavior of the device, I imagined that the fresh telnet connection gives us the latest data and there is no buffering issue.

So I tweaked the data logging code to establish the telnet connection every time the values are asked. The connection is closed after the every data acquisition. I like this as we can also make the test connection between each data acquisition points, although I have not tried it yet. The code is in the same folder named ctc100_controller_v2.py

5. Heating

Now I thought that I did all I wanted to do this evening, so the heater was turned on at ~20:50, Jul 21. The heating power saturated at 22W, which is the set limit.

- Temperature Log updated 2021/7/23 12:00 Heating Ended

- Assuming reaching the room temp at ~30hrs and heating power saturated at 22W: Predicted heat injection 30*3600*22 = ~2.4MJ

Update from Stephen
- Note that we can check logging accuracy against the snapshot (timestamp 20210723_1113).
If my math is correct, this would be time = 37.35 38.35 hours

Update from KA
=> The corresponding time in sec is 138060 sec
The raw data line for the corresponding time is:

138016.839614, 295.805, 306.678, 302.518, 312.401, 0.000, 0.000, -0.001, 0.621, 0.622, 1.429, 0, 0, NaN, NaN, NaN
The values on the photo 295.806, 306.677, 302.518, 312.401 ==> Well matched. Victory!

[Returned] Brought one HAM-A coil driver (D1100687 / S2100619) and one Satellite Amplifier (D1002818 / S2100741) from the 40m

Also brought some power cables.

Brought ~1m of 0.0017" (~43um) misical wire. This will make the tension stress be 341MPa. The safety factor will be ~7.

The following radiation cooling model well explained the cooling curve of the test mass (until ~150K)

where dQ/dt is the heat removed from the test mass, A is the surface area of the test mass, is the Stefan-Boltzmann constant, T_SH and T_TM are the temperatures of the surrounding shield and the test mass.

Can we extract any information from this "0.15"?

I borrowed "Cryogenic Heat Transfer (2nd Ed)" by Randall F. Barron and Gregory F. Nellis (2016) from the library.
P.442 Section 8.5 Radiant Exchange between Two Gray Surfaces can be expressed by Eq 8.44

where T_i is the temperature of objects 1 and 2. For us, OBJ1 is the test mass and OBJ2 is the shield. A1 is the surface area of A1. F_1,2 is the view factor and is unity if all the heat from the OBJ1 hits OBJ2. (It is the case for us.)

is an emissivity factor.

The book explains some simple cases in P 443:

Case (a): If OBJ2 is much larger than OBJ1, where the e_i is the emissivity of OBJi. This means that the radiated heat from OBJ1 is absorbed or reflected by OBJ2. But this reflected heat does not come back to OBJ1. Therefore the radiative heat transfer does not depend on the emissivity of OBJ2.

Case (b): If OBJ1 and OBJ2 has the same area, . The situation is symmetric and the emissivity factor is influenced by the worse emissivity between e1 and e2. (Understandable)

Case (c): For general surface are ratio,  . OBJ2 receives the heat from OBJ1 and reradiates it. But only a part of the heat comes back to OBJ1. So the effect of e2 is diluted.

For our case, OBJ1 is the Si mass with DxH = 4in x 4in, while the shield is DxH = 444mm x 192mm. A1/A2 = 0.12.
We can solve this formula to be Fe=0.15. e1 = (0.147 e1)/(e2-0.0178).

Our inner shield has a matte aluminum surface and is expected to have an emissivity of ~0.07. This yields the emissivity of the Si test mass to be e1~0.2

How about the sensitivity of e1 on e2? d(e1)/ d(e2) = -0.95 (@e2=0.07).

Can Aquadag increase the radiative heat transfer?

Depending on the source, the emissivity of Aquadag varies from 0.5 to 1.
e.g. https://www.infrared-thermography.com/material-1.htm / https://www.mdpi.com/1996-1944/12/5/696/htm

• Assuming Aquadag's emissivity is ~1
  • If only the test mass is painted, F_e increases from 0.15 to 0.39 (x2.6)
  • If the inner shield is also painted, F_e increases to 1, of course. (pure black body heat transfer)
  • If shield panels are placed near the test mass with the inner surface painted, again F_e is 1.
• Assuming Aquadag's emissivity is ~0.5
  • If only the test mass is painted, F_e increases from 0.15 to 0.278
  • If the inner shield is also painted, F_e increases to 0.47.
  • If shield panels are placed near the test mass with the inner surface painted, F_e is 0.33 assuming the area ratio between the test mass and the shield panels to be unity.

It seems that painting Aquadag to the test mass is a fast, cheap, and good try.

Updated Jul 26, 2022 - 22:00

1. Reconstruct the cryostat
   1. [Done] Reinstall the cryo shields and the table (Better conductivity between the inner shield and the table)
   2. [Done] Reattach the RTDs (Inner Shield, Outer Shield)
      -> It'd be nice to have intermediate connectors (how about MIllMax spring loaded connectors? https://www.mill-max.com/)
   3. Reattach the RTD for the test mass
2. Test mass & Suspension
   1. [Done] Test mass Aquadag painting (How messy is it? Is removal easy? All the surface? [QIL ELOG 2619]
   2. [Done] Suspension geometry change (Higher clamping point / narrower loop distance / narrower top wire clamp distance -> Lower Pend/Yaw/Pitch resonant freq)
   3. [Done] Setting up the suspension wires [QIL ELOG 2620]
   4. [Done] Suspend the mass
3. Electronics (KA)
   1. [Done] Coil Driver / Sat Amp (Power Cable / Signal Cables)
   2. Circuit TF / Current Mon
   3. [Done] DAC wiring
   4. [Done] Damping loop
4. Sensors & Calibration (KA)
   1. [Done] Check OSEM function
   2. [Done] Check Oplev again
   3. Check Oplev calibration
   4. [Done] Check Coil calibration
   5. Use of lens to increase the oplev range
   6. Recalibrate the oplev
5. DAQ setup (KA)
   1. [Done] For continuous monitoring of OSEM/OPLEV

[Stephen Koji]

We decided to paint the silicon test mass with Aquadag to increase the emissivity of the test mass.

Stephen brought the Aquadag kit from Downs (ref. C2100169) (Attachment 1)

It's a black emulsion with viscosity like peanut butter. It is messy and smells like squid (Ammonium I think) (Attachment 2)

We first tried a scoop of Aquadag + 10 scoops of water. But this was too thin and was repelled easily by a Si wafer.
So we tried a thicker solution: a scoop of Aquadag + 4 scoops of water. (Attachment 3)

The thicker solution nicely stayed on the Si wafer (Attachment 4)

We want to leave the central area of the barrel unpainted so that we can put the suspension wire there without producing carbon powder. (Attachment 5)
1.5" from the edge were going to be painted. The central1" were masked.

The picture shows how the Si test mass was painted. The test mass was on a V-shaped part brought from the OMC lab. The faces were also painted leaving the mirror, while the place for RTD, and the magnet were not painted. (Attachment 6)

It looked messy while the painting was going, but once it started to dry, the coating looks smooth. It's not completely black, but graphite gray. (Attachment 7)

After the test mass got dry, another layer was added. (Attachment 8)

Then made it completely dry. Now the mask was removed. Nice! (Attachments 9/10)

[Stephen Koji]

While Stephen worked on the RTD reattachment, I worked on the suspension part.

- First of all, we found that the magnet was delaminated from the silicon mass (Attachment 1). It was bonded on the test mass again.

- The suspension frame was tweaked so that we have ~max suspension length allowed.

- The first attempt of suspending the mass with steel wires (0.0017" = 43um dia.) failed. Stephen and I went to downs and brought some reels.

- I chose the wire with a diameter of 0.0047" (= 119um) (Attachment 2). ~8x stronger! The suspension was successfully built and the mass is nicely sitting on the 4 strain releasing bars (improvised effort). (Attachments 3/4)

We can install the suspension in the chamber tomorrow (today, Wed)!

[Stephen Koji]

Road to cooling down

• The suspension with the test mass was installed in the chamber again
• Looking at the oplev beam, we jiggled the wire loop position to adjust the alignment approximately.
• The oplev beam was aligned more precisely.
   
• We intentionally kept the OSEM at the "fully-open" position, while it is still close to the magnet so that we can have some actuation.
• The coil driver was tested before closing the chamber, but it did not work.
  The coil itself was still intact, and the mirror was responding to the coil current if the coil current of ~100mA was applied from a bench power supply with the current ~100mA).
  So the problem was determined to be external.
   
• Once we were satisfied with the oplev/OSEM conditions, the inner and outer lids were closed. Then the chamber was closed.
   
•  Started pump down.
• Started cooling down @18:30 / started temp logging too. Log filename: temp_log_cool_down_20210728_1830.txt

The photos were uploaded to Google Photo of WB labs.

The coil driver issue was resolved:

• It was necessary to take care of the enable switch. Made a DB9 short plug for this purpose.
• The output R was 1.2K (i.e. 2.4K across the + and - outputs). We needed ~10x more to see visible motion of the mass
• e.g. The internal gain of the driver is x1.1. If we connect 5VDC input across the diff input of the driver yields, +11V shows up across the outputs of the final stage.
  If the R across the coil is ~100Ohm, we get ~100mA.
• Soldered 6 x  330Ohm (1/8W) in parallel to 1.2K R_out. -> This ended up 51.5Ohm x2 across the coil. Each R=330 consumes ~1/10W. ->OK

Checking the DAQ setup / damping loop

• DAQ setup
  • ADC: QPD X->FM16 / Y->FM17 / S->FM18 / OSEM-> FM19
  • DAC: CH11 -> Coil Driver In
• Connected FM16 and FM17 to the coil drive by setting C4:TST-cdsMuxMatrix_12_17 and C4:TST-cdsMuxMatrix_12_18 to be 1.0
• It was not obvious if the coil could damp the rigid body modes.
  • Actating the magnet caused Yaw motion most. Some Pitch motion too.
  • Configured FM16 and FM17 for the damping loop.
    • Filter Bank #1: [Diff0.1-10]  Zero 0.1Hz / Pole 10Hz
    • Filter Bank #10: [Anti Dewht]  Zero 1&200Hz / Pole 10&20Hz
  • Tried various damping gain. The mass was moving too much and the proper gain for the damping was not obvious.
  • So, the initial damping was obtained by shorting the coil at the coil in of the sat amp unit. (Induced current damping)
  • Once the test mas got quieter, it was found that -0.01 for FM16 could damp the yaw mode. Also it was found that +0.1 for FM17 could damp the pitch mode. (But not at once as the filters were not set properly)
     
• TF measurement for calibration
  • The beam was aligned to the QPD
  • The test mass was damped by using the damping loops alternately 
  • Taken a swept sine measurement Filename: OSEM_TF_210729_0243.xml
    Recorded the time, saved the data, and took a screenshot
    • This measurement was taken @T_IS=252K / T_TM=268K @t=8hr (2:30AM), Rcoil=15.6Ohm
  • Second measurement Filename: OSEM_TF_210729_2147.xml
    • @T_IS=172K / T_TM = 201K @t=27.5hr (10PM), Rcoil=10Ohm
  • 3rd measurement Filename: OSEM_TF_210730_1733.xml
    • @T_IS=116K / T_TM = 161K @t=47hr (5:30PM), Rcoil=?
  • 4th measurement Filename: OSEM_TF_210731_2052.xml
    • @T_IS=72K / T_TM = 134K @t=75hr (9:30PM), Rcoil=6.0Ohm

OSEM LED/PD

• The Satellite amp brought from the 40m is used as-is.
• The initial OSEM reading was 8.8V, this corresponds to ~30000cnt.
• As the OSEM was cooled, this number was increasing. To avoid the saturation, a voltage divider made of 4x 15kOhm was attached. I didn't expect to have the input impedance of the AA filter (10K each for the diff inputs), this voltage divider actually made 18.24V across POS and NEG output to be 5.212V to the AA fiter. So the voltage division gain is not 0.5 but 0.2859.
• This made the ADC range saved, but we still have a risk of saturating the PD out. If this happens. The PD TIA gain will be reduced before warming up.
  -> The TIA and whitening stages use AD822, and the diff output stage uses AD8672. AD822 can drive almost close to rail-to-rail. AD8672 can drive upto ~+/-14V.

There was not enough time for the QPD calib -> Tomorrow

The current cooling curve suggests that the radiative cooling factor Fe (black body =1) increased from 0.15 to 0.5.

Update: The test mass temp is reaching 200K at ~27hrs. cf previously it took 50hrs
Update: The test mass temp is 170K at ~41.5hrs.

OSEM illumination & photodetector efficiency has been kept increasing @41.5hrs

In all aspects, the latest cooling shows the best performance thanks to better thermal connection, thermal isolation, and the black paint.

- The cold head cooling is faster and cooler

- The inner shield cooling is faster

- The test mass cooling is faster

The test mass temperature indicates 121K@100hr but there seemed a few sensor glitches for the test mass (𝛥=-4.2K) and the inner shield (𝛥=-0.43K).
So the actual test mass temperature could be 125K.

The temp was read to be 119K@114hr (Attachment 1)

There was very little cooling capability left for the test mass (Attachment 2)

The OSEM reading is now stable @12.3V (Attachment 3)

The raw temp data and the minimal plotting code are attached (Attachment 4)

[Stephen Koji]

We applied Aquadag painting on the inner side of the inner shield.

• Upon the painting work, we discussed which surfaces to be painted. Basically, the surface treatment needs to be determined not by the objects but by the thermal link between the objects.
  • We want to maximize the heat extraction from the test mass. This means that we want to maximize the emissivity factor between the test mass and the inner shield.
  • Therefore the inner barrel surface of the inner shield was decided to be painted. The test mass was painted in the previous test.
  • For the same reason, the lid of the inner shield was painted.
  • It is better to paint the cold plate (table) too. But we were afraid of making it too messy. We decided to place the painted Al foil pieces on the table.
     
  • The outer surface of the inner shield and the inner surface of the outer shield: Our outer shield is sort of isolated from the cold head and the steady-state temp is ~240K. Therefore we believe that what we want is isolation between the inner and outer shields. Therefore we didn't paint these surfaces. (note that in Mariner and beyond, the outer shield will be cooled, not isolated, and the radiative link to the outer shield would be strong by design)
  • I believe that this is not the ideal condition for the inner shield. We need to model the cryo stat heat load and take a balance between the isolation and the conduction between the outer shield and the cold head so that we gain the benefit of the outer shield as a "not so hot" enclosure.
     
• OK, so we painted the inner barrel of the inner shield, the lid of the inner shield, and some Al foils (shiny side).
• Stephen made the Aquadag solution. The solution was 2 scoops of Aquadag concentrate + 6 scoops of water, and the adhesion/runniness test was done on a piece of aluminum foil.
• The barrel and the lid were painted twice. Attachment 1 shows the painting of the inner shield cylinder. Attachment 2 shows a typical blemish which necessitates the second coat.
• To accelerate the drying process, we brought the heat gun from the EE shop --> (update - returned to EE shop, see Attachment 3)
   
• We took some photos of the process. They are all dumped in the QIL Cryo Vacuum Chamber Photo Dump album in the ligo.wbridge account.