Aidan and I are acquiring a beat signal from the interference of a reference beam and a frequency-modulated beam. The demodulated output can be seen on THIS ENTRY.
I have attached a photograph of our setup below. The reference beam is a pickoff from the 1W NPRO, and the zeroth order AOM beam is blocked by an iris.
Aidan will shortly post a schematic of the electronics, and we hope to start actuating on the laser frequency later today.
I have rebooted the network. All internal IP addresses seem to be pingable but cdsrana wifi is returning garbage self assigned IP address. There is also no connection to the outside internet.
I plugged my laptop directly into the LIGO network port and DHCP failed, my computer is assigning itself and IP and so nothing is accessable to the outside world. I also tried pinging the static IP address accociated with out lab, no response from our router.
I've emailed Larry Wallace to see what is up. This is not something I can resolve if something is wrong with the LIGO network.
Larry Wallace was still on campus and came around. Apparently he wants to replace the switch in the entrance to the ATF anyway with something that isn't 20 years old.
It turns out the issue was a combination of a poor connector from the LIGO network switch to the wall socket (wobbly patch) AND a bad cable from the wall socket to our gateway router 10.0.1.1. The LIGO switch patch is now pushed firmly in and the offending wall-to-our-router cable is in the bin. Fault was a broken clip on the cable. Essential network links should be firmly clipped in.
There were also some config changes to both the 10.0.1.1 gateway and 10.0.1.2 wireless bridge router. I have screen shot the new configuration and put it on the ATF wiki page. It seems like the previous configuration was to feed the network connection through regular connection ports (rather that "internet" input) on the gateway router (10.0.1.1).
Internet is now accessible through the local ATF network and into the PSL lab computers
I power cycled the Linksys router but that didn't do anything. Then I power cycled the 3com switch, and it came back on again.
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.
Inverted Pendulum links:
I estimated what the inverted pendulum leg's spring constant (K) should be using some realistic/desired parameters.
Me and Kate assumed the inverted pendolum made of a cylindrical steel leg of radius 0.5 cm and lenght 0.42 m, and a point mass of 1 kg. With a resonant frequency of 40 mHz we obtain:
I talked with Steve Penn at GWADW, and he made a suggestion for an inverted pendulum design based on something he built some time ago to measure the thermal noise of fused silica fibers. His sketch is attached and I've labeled key parts. The sketch on the left provides a side view and the sketch on the right is a top-down view. The thick rod-like ends of a pulled fiber are welded to glass disks which can be easily clamped to a metal structure. The clamps are not depicted below, but would take the form of an 'L', sandwiching the glass disk against the metal plate. The clamps themselves would be screwed to the plate. The top plate is attached to support arms which are fixed rigidly to a table (or, in our case, to the rhomboid). The structure beneath the glass fiber has three legs that extend up and beyond the support arms. At the top of these legs is a disk and on top of the disk is a vertical rod with a mass that can slide up and down. The height of the mass is adjusted to set the resonance frequency.
This seems like a design that is quite compatible with what we need. The moveable mass would need to be designed such that we can fix a mirror for the Michelson on top of it. Steve also suggested that if we want to limit the number of degrees of freedom, we could pull a glass ribbon instead of using multiple fibers.
I have included below my proposal for the design of the Iodine-stabilized laser system. It is a slight variation on the standard implementation of the FM spectroscopic approach to frequency stabilization, and is sort of a hybrid between the two designs investigated by Leonhardt and Camp (article).
Brief overview of my implementation of laser frequency stabilization by FM spectroscopy:
1. Nd:YAG laser outputs ~700mW infrared (1064nm) light, which is passed through a Faraday isolator and the frequency doubling oven (stabilized by Zach's precision temperature controller)
2. Dichroic mirror transmits second harmonic, reflects the infrared light for optical heterodyne detection with IR output from second identical laser system (each system locked to a difference resonance)
3. The 532nm light is split into pump and probe beams, whose relative intensity can be adjusted to optimize the SNR (minimize power broadening, maximize signal strength)
4. Probe beam is RF phase modulated by the EOM, adding a pair of weak FM sidebands to the 532nm carrier signal.
5. Probe is then passed through Iodine vapor cell, overlapping temporarily with the saturating pump beam. The partial depletion of the Iodine ground state by the pump beam causes each frequency component of the probe beam to experience different levels of absorption and dispersion.
6. Probe impinges on a square-law photodetector, which converts the time-varying field amplitude of the probe beam into a beat signal at the RF modulation frequency. This RF heterodyne beat signal contains the imprint of both the absorption and dispersion profiles of the spectral feature of interest.
7. The beat signal is then fed into a double-balanced mixer (DBM) along with the local RF oscillator which drove the EOM. Assuming the beat signal has been appropriately phase-matched with the oscillator, the output from the DBM should be the dispersion profile of this given spectral feature at DC, which is an ideal error signal for frequency locking.
8. The error signal is then fed back into the laser head controller where the fast and slow internal frequency actuators work to lock the laser.
Any questions and/or feedback regarding this layout would be appreciated. This overview was very rough, and clarification can be provided as needed. More details will follow as soon as I begin rounding up the necessary components in lab tomorrow.
As Zach observed in a previous post (elog #1560), the BIPM Iodine cells he shipped from Goddard arrived with globs of Iodine scattered randomly throughout the cell. Before we could put them to use, we needed to find some way to localize the globs of Iodine into the cold fingers. The general idea was to use some heating element to gently melt the solid chunks of Iodine and carefully direct the liquid into the cold finger.
Well, I gave this a try today with the heat gun Zach left in the lab for me, and it was a bit more difficult than I imagined it would be. The heat gun reached its maximum temperature rather quickly and has no temperature control settings beside "hot" and "cold", so I altered the cell temperature by placing the heat gun on its flat back mount and varied the cell height above it. With the heat gun at max temperature, I observed the following:
Iodine vapor and liquid are supposed to have distinct colors, but when present in such miniscule amounts I couldn't see either of them. The best way I found to move the globs into the cell finger was to hold the cell more than 30cm away from the heat gun and try to slide them into the finger. It took a bit of practice, but I think as a first (and very rough) attempt, it was pretty successful.
Below is a picture of the first of two BIPM cells I modified today. The picture was taken when the cell had been cooling down for ~5 minutes, so you can distinctly see both the solid and liquid phases of Iodine at the bottom of the finger. Those dark colored specs used to be scattered randomly throughout the cell and after my manipulations today I couldn't find any of them outside the finger. I've also heard the color of liquid Iodine described as "orange-brown", so I assume the ring above the solid Iodine specs is condensed Iodine vapor still cooling down. Unfortunately, I had to leave shortly after making these manipulations, but I look forward to seeing how they look after completely cooling down to room temperature. As a final touch, once all the Iodine present in the finger has solidified, I will slightly heat it while holding the cell with the cold finger pointing downward in an attempt to make all the scattered globs merge into a single clump of Iodine.
I am attaching the datasets for the phase and frequency noise plots for the Isomet and the Marconi. The structs also include the noise measurements taken from the spectrometer and the input amp in both phase and frequency space. As soon as I get the NEOS measurement working, I will upload some plots for that and the data sets as well.
Here is the fft of the output of the Isomet oscillator (10mW output power, with a 50ohm attenuator on the frequency modulation input). The input of the spectrum analyser had the input attenuator set to 20dB.
I centered the measurement around the oscillator frequency (approx 30MHz). We see large peaks every 132kHz and smaller sidebands on each of these.
Although this is bad, I'm not sure that this is the issue. We'll have to diagnose it on the bench, but I think it just may be the way you guys were powering it.
Those lines are probably from the switching power supply; the type of RF amplifier that's in there probably has no power supply rejection at all. We'd be
much better off powering this thing from a real, regulated DC power source and then make sure to put the bypass capacitors (one big, one small) right on the power pins of the AOM driver.
Agreed we didn't have the capacitors right up against the AOM driver. They were on the power supply, and the large ones could be bigger.
However we did try it with a second totally different style of power supply, and the frequency of the lines was exactly the same. It will be interesting to see whether a regulated supply gets rid of them.
We have large glass jugs of methanol and acetone - we need one of isopropyl too. I'm not sure how to go about this.
One in the TCS Lab. One by the ATF door. I have more ant bait available in the TCS Lab.
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.
Okay - all the steps in the procedure of eLOG 2476 have been verified as working - with the exception of the RTDs in the chamber.
With regards to taking dark noise spectra at different biases and temperatures, looks like Raymond took spectra with biases of [50, 100, 200, 400, 600, 1000]mV. If no objections, I’ll stick to that number of measurements.
I’m a bit pushed for time with other stuff. I wonder if the shield RTD is sufficient to run tests on the system? I’ll go back through the data and see how reproducible the relationship between shield temperature and PD temperature is. If it is reliable then in the interests of time, I’m going to forgo re-installing the extra RTDs in the chamber just now.
Looks like the temperature difference between the PD and the shield is relatively small. Even the transients when the heater is applied are order 5K.
This means that, for quick purposes, the shield RTD is a good proxy for the PD temperature.
The attached data is the difference between PD and shield RTD from circa 5th-6th February 2020.
I recorded a 15 minute overview that describes the JPL PD set up and how to operate it. I'm in the process of embellishing the operation procedure (previous version can be found here: eLOG 2476).
[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 keys to the 35W (as well as the gyro laser) have been removed. There is no elog entry saying who did this, where they are, or when it happened.
[Edit: We found it in one of the drawers of the console tables!]
- for personal use only -
I've returned the Keithley Source Meter unit
- The unit (Keithley 2450?2460?)
- A power cable
- A pair of banana clips
- the transistor test fixture & triax cable/connectors
Note that the back panel connectors are Triax, not the usual Coax.
I added some EPICS channels to the Hartmann sensor softIoc and then added these to be recorded in the frames.
I then killed the daqd process on fb1 so it would start afresh.
I killed and restarted the daqd process because I wanted to add some 16Hz TCS channels to the frame builder. These are from the Athena DAQ box.
I edited the following files:
We are continuing the naming scheme with CX? I thought we were planning on making C2 the whole subbasement...
Nope. Each lab has its own number. I think PSL is C3, TCS is C4, ATF gets C2....Frank has the full listing of these things. Or at least that's the latest I heard a few months ago after several elogs back and forth.
Right - I am questioning the scalability and sense of this scheme, and inquiring if this is aesthetic.... i.e. is there a reason it cannot be C2SYS for all of bridge, (possibly front end naming limitations)?
I have run the design rule check on the new PCB and there are no violations. I am going to do one last run of checking dimensions (hole placement, BNC mount clearance, etc.) and then generate files. Once I have them, provided there are no more comments, I will take them to Steve or someone to purchase.
I thought of a primitive device that could be used as a frequency discriminator in our (and similar) readouts. It is sort of like the "cable" method championed earlier this year, only instead of using an asymmetric michelson style setup for the phase shift, one uses a high-Q LC filter tuned to the carrier in one path, as below:
As the signal from the PD changes frequency, the phase shift from the LC filter changes, providing a linear output from the mixer proportional to the frequency shift. Using high-enough-quality components (DCR ~ 50 mOhm), we can get discriminants approaching 1 mrad/Hz, which would require hundreds of km of cable in the other case.
Feeding to the LO with the LC leg ensures that AM from the filter---which is already higher than first-order---is removed.
There are two probable issues:
I talked to Frank before thinking about this much further, and he didn't think there was anything obviously wrong with the idea. To the contrary, he thought it might be worth looking at because it could be useful for other experiments. I don't think we have good enough components to try this yet. We may have some good capacitors somewhere, and Frank mentioned that we might have a DIY inductor kit, so that's promising. Otherwise, I think we might want to order some good high-Q components and try this out.
The reason I went for cable length instead of LC, is that the LC is basically a version of a VCO with a PLL, except that the VCO is made of macroscopic L & C. I think the only way to be
sure about the performance is just to build it and see. But basically, you can be sure that you have to ovenize the LC and pack it with putty. Might as well get yourself warmed up for some
ovenizing one way or another: gyro, table, EOM, LC, etc.
Yes, I can see that. I also figured it was because the intended purpose of the cable method was to measure much larger frequency shifts, not mHz-level. So they shouldn't really be compared directly.
I'm building a simple LC to see how it performs. If it does even close to well then maybe I'll order some nicer parts.
Over the past couple days (in between overhauling malfunctioning PDs), we have been trying to hunt down the excess low-frequency noise. Last week, with Koji's help, we essentially ruled out the possibility of scattered light noise from the transmitted end of the gyro by intentionally reflecting varying amounts of light back into the cavity and observing NO DIFFERENCE in the low-frequency spectrum.
AM from the EOM still seems to be the most likely suspect, and we continue to find ways in which it could couple. Yesterday, we think we traced the crazy AM level modulation that I mentioned in my last post to the iris we are using to isolate the proper beam out of the AOM setup. The double-1st order (desired) and single-1st order (undesired) beams are extremely difficult to isolate from one another, so we had to use a very small aperture that we think was coupling EOM jitter into AM quite strongly. We moved the iris down the beam path closer to the waist(s) and were able to get better isolation with a larger aperture. The drift in AM level now seems absent.
While the AM levels in the CCW and CW beams (as measured on the RF analyzer) are not simultaneously minimized at the same pre-EOM HWP angle, there is no longer a large-angle discrepancy; the difference is <1 degree and the noise in both beams can now be kept fairly low at some compromise angle. I have set up a PDA255 in each path immediately before the cavity to monitor them both simultaneously. By adjusting the HWP before the EOM and fine-tuning the EOM orientation, I was able to get the AM peaks in both beams to be <10 ppm relative to the carrier.
After re-locking the gyro, I saw NO IMPROVEMENT in the low-frequency noise yet again. I replaced the PDA255s and tried looking at the LF noise directly after demodulation. I saw some excess noise above the broadband floor below about 1 Hz, which is roughly where the excess gyro noise begins. Upon deliberately de-tuning the HWP so that the RF peaks were >20x higher, however, I saw NO NOTICEABLE CHANGE in the audio spectrum. Looking at the peaks in the RF spectrum more closely, I discovered that the linewidths are below 1 Hz, suggesting that this may not be the source after all.
I suggest that we vent the chamber and obstruct the cavity so that we can run the same tests with the REFL PDs themselves. Then, we can use the noise we measure in the error signal to easily and accurately estimate the contribution to the NB.
The low frequency noise is still present in the TRANS signal, seemingly discounting the beam jitter theory. I think I am too sad to ask Rana for the $5, though.
I am still absolutely stumped about the LF noise. Today I revisited and reconfirmed the fact that the gyro noise is strongly correlated with the noise in the DC_TRANS level at low frequencies:
What this means remains a mystery to me. I also re-verified that the laser power fluctuations do not have the right shape to explain this noise (nor is there an obvious coupling from laser intensity noise to gyro noise in this case). The fact that we have ruled out traditional RAM as the source makes it anyone's guess as to what's happening here. Increasing the LF common-mode gain by ~50 has zero effect on either the gyro noise or the DC_TRANS noise.
I am pointing out the DC_TRANS correlation again because I think it is a potential smoking gun. If you or someone you know has any information regarding this matter please call 1-800-GYRO-WTF.
I replaced the LT1128 in the whitening filter for channel B with one from Downs, to see if it exhibited better low-frequency current noise. It did, but only very marginally:
The Downs part is about a factor of sqrt(2) better, but even it is not appreciably below the "max" estimate from LISO. However, the fact that the noise improves slightly here (and nowhere else) when that one LT1128 is replaced only strengthens my suspicion that this is indeed just bad current noise performance from the LT1128.
In fact, if you look at the performance of the aLIGO in-vacuum DCPD preamps, which have essentially same LT1128 whitening stages as I used in the M2 (actually 2 in series), you see a similar amount of excess low-frequency noise. Here is a measurement of one DCPD I made while at LLO, compared with LISO estimates using the "typical" and "maximum" noise specs for the LT1128:
As you can see, the data match the "max" curve very well. So, either a.) Linear is pulling a fast one on us and under-reporting the typical current noise or b.) we are doing something silly with the circuit, and this is somehow perfectly mimicking a "max" current noise part.
For a while now I've wondered how LTSpice thinks the the AD829 should perform around 1 MHz. To that end, I ran simulations of an AD829 inverter for a few different gain/compensation settings.
This is also of potential interest to the 40m and cryo, both of whom may want to construct detailed phase budgets in the region from 100 kHz to a few megahertz.
As a reminder, this is how Analog Devices thinks we should stuff the AD829:
What I simulated was slightly different:
Here R3 means I've added a resistor from the (+) pin to gnd, in order to cancel some of the bias current.
For each gain, I ran the simulation with both a 50 Ω load and a 1 MΩ load. In the attached zip file you can find screenshots of the Spice model for each gain setting.
Visitors to the lab ...
A 121307 161942 0100 0.3 000000 0.5 000000 TMP 000680 R/H 000130 LOC 000000 C/S 000E54
A 121307 164641 0100 0.3 000000 0.5 000000 TMP 000680 R/H 000130 LOC 000000 C/S 000E53
A 121307 171340 0100 0.3 000000 0.5 000000 TMP 000685 R/H 000130 LOC 000000 C/S 000E52
A 121307 174039 0100 0.3 000000 0.5 000000 TMP 000680 R/H 000200 LOC 000000 C/S 000E53
A 121307 180738 0100 0.3 000000 0.5 000000 TMP 000680 R/H 000210 LOC 000000 C/S 000E57
A 121307 183437 0100 0.3 000000 0.5 000000 TMP 000680 R/H 000210 LOC 000000 C/S 000E56
A 121307 190136 0100 0.3 000000 0.5 000000 TMP 000680 R/H 000195 LOC 000000 C/S 000E5C
A 121307 192835 0100 0.3 000000 0.5 000000 TMP 000685 R/H 000195 LOC 000000 C/S 000E69