Mitt liv som Tobin

Also, another. This one has our actual datastream on the audio track instead of CCR. You know that scene in Contact where Jodi Foster is listening to the radio telescope signal on headphones? It's kind of like that. We listen.

Awesome photos from CERN:

http://www.boston.com/bigpicture/2008/08/the_large_hadron_collider.html

Suppose we have an opamp connected in the inverting amplifier configuration:

 in        R1                      R2       
 o----+--/\/\/\------------+-----/\/\/\-------------+----o out
                           |                        |
                           +--|-\                   |
                              |  \__________________|
             R_BIAS           |  /
      +------/\/\/\-----------|+/ OPAMP
      |
     GND

The gain of this thing is just -R2/R1, i.e., if R2 is 10000 ohms and R1 is 5000 ohms, then the output will be twice as large as the input, and with its sign flipped. That is well and good, but we are concerned about the noise introduced to the circuit. Noise is introduced both by the opamp itself, which, in our abstract view, is intrinsic to the amplifier; and by Johnson noise in the two resistors in the feedback network.

Horowitz and Hill give an expression for the noise of this circuit on page 447 of Art of Electronics, but they dedicate remarkably little space to it. Only about 8 3/4 square inches, actually, and their answer is somewhat confusing.

First, the amplifier noise. The amplifier noise is specified as an equivalent voltage noise e_a and current noise i_a seen at one (or both, you never can be sure what the datasheet is quoting) of the opamp terminals. We can pretend that the voltage noise e_a is all clumped at the noninverting input to the opamp. We then multiply that by the "noise gain," which is the transfer function from the opamp's noninverting input to the output of the circuit: 1+R2/R1. Done.

One of kennyjensen's grad school projects is slashdotted today: "Single carbon nanotube is fully functional radio, receiving music over standard radio bandwidth".

Last night I went to a hot tub party at Kip Thorne's house. Stephen Hawking showed up. No joke!

He didn't go into the hot tub, though.

Currently at Ontario airport. 6:15 AM flight = UGGGGHH. Baton Rouge, here I come.

Flying turned out to be no trouble at all, and I even got a complimentary in-flight mojito! Also, I realized I was quite dressed the nerd part, with my One Quark Two Quark Red Quark Blue Quark T-shirt and reading Schultz's A First Course in General Relativity, both augmenting the pre-existing frazzled-gradstudent look. But, let me remind you, a nerd drinking a complimentary in-flight mojito.

( comments on Schultz's book )

1. Bicycled down south plymouth avenue to north plymouth avenue.. Tucked into church street to visit City Hall for the Neighborhood Associations Presidents' Meeting. I hadn't even known where it was located. It's a smaller place than you think, but in grand style--what I'd call "Upstate Gothic," if that makes any sense at all. I locked up my bike to the flagpole bearing the Rochester flag. It was all very anti-climactic, though. First I couldn't figure out how to get in the building (the front door is locked after hours). And then it turned out that the meeting had been mysteriously cancelled. I guess I'll have to wait a little longer to meet John Borek.

2. Having some free time, I biked out State to check out Flat Iron Cafe, which is located where State, Lake, Smith, and Lyell streets and avenues come together. There were just two guys there, a well-dressed middle-aged man and an older man sitting at a wrought-iron table outside, enjoying espresso. The younger one, Tom, turned out to be the owner; the older one spoke with a thick Italian accent. I sat and talked with them for a while. It's a good place— I plan to return one night with a backgammon set and my officemate Stefanos. Some review I read of the place suggested it as a place to read Kerouac. I read Spivak's Calculus on Manifolds.

At one point while there, cops suddenly descended on the place. Three officers on foot from one side, and a car with lights from the other. But the car with lights was a total coincidence, and the cops on foot were there as customers.

Tom and partners are also opening up a comic book store behind the cafe.

3. Riding home through downtown, I chatted with hypostatization on the phone, he currently on the road in South Carolina. At some point out on West Main, a bunch of kids approached me. "Get off that bicycle!" one of them yelled, apparently wanting to steal it. I've been approached by kids maybe two or three times this summer and threatened in some way, but this was the only time it was at all serious. Still—I never know how serious these would-be muggers are and my first response to muggings is apparently indignation. "I don't think so!" I yelled back. "Give me that bike, I'm serious!" "What the hell?" It occurs to me that I should flee—I know this kid could beat me up pretty bad, and there's a half-dozen of his cronies closing in. I ride out into traffic. He runs after and grabs my backpack. Some older black man leans out of an SUV and yells "Don't do it, man!" and the kid lets go briefly, I ride off.

A couple blocks later, I called Rob back.

"Was that a mugging?" he asked.

Then out in South Carolina there was something like a gunshot and he had to go. Weird night, huh?

4. Back at the co-op we're moving into an exciting transitionary phase—Summer residents Mike, Leland, Lewis, Jon McVay, and Andrew are leaving, moving back into the dorms. And Jon P., Nicole, Kastan, Bree, and Zach are moving into the house. Doug, Becky, and Jeremy are moving into an additional house across the street. Exciting times. We're going to have an orientation of some kind next weekend. Perhaps to involve wine and cheese. My twister suggestion may have been vetoed.

5. Critical Mass tomorrow!

After several days of fighting with the peculiarities of the computer system at LHO (LIGO Hanford Observatory), my analysis program is good to go—and it is awesome indeed. All I need to do is type "make submit" and my program finds the data to be analysed, constructs a tree containing thousands of processing jobs that will perform the analysis, and then submits these jobs to the computing cluster at LHO. Working late at night on a Sunday when nobody else is using the cluster, I have all of the computing resources to myself: 276 processors, each running at several gigahertz, each one with a gigabyte of RAM, with a collective total of twenty seven terabytes of hard disk space, and access to a robotic magnetic tape system containing who knows how many terabytes more. Amazingly all of this sits in one room in a remote, desert corner of Washington State. Amusingly the data processing program consists of only 216 lines of Matlab, 100 lines of Python, 26 lines of Makefile, with a pinch of sed and dash of other unix goodness thrown in for good measure. Programs that write programs: that's what makes CS fun. As I said: type "make submit", wait, then results.mat will be sitting there, ready for ingestion by Matlab and consumption by the analysis phase.

Kent and I joked that a proper scientific analysis program should start with the raw data and ultimately typeset and mail the resulting publication. I wasn't really kidding. Read paragraph three of Background: Kepler. This time I might have it: type "make analysis" and that results.mat file will be picked up and churned through the analysis. Hypothesis testing is done. Figures are generated, in PNG and EPS. The LaTeX typesetter is run, producing both postscript and PDF. If I were properly awesome, the proper concluding text will be subbed in given the results of the hypothesis testing. Ultimately: Type "make all." Robots in Washington State awaken, transcribe signals from magnetic tape to disk. An army of machines grinds through the data. A program in New York State performs model fitting, hypothesis testing. LaTeX turns this into a beautiful manuscript, which the Laserjet 2300dn fuses to the page.

Pull the page out of the printer and read what you have discovered.

IT's been a busy week at LIGO for me.. taking data, installing electronics, soldering, assembling, testing, spectrum-analysing, programming, etc. The LIGO electronics shop here has all the oscilloscopes, multimeters, spectrum analysers, function generators, power supplies, cables and connectors of every sort, etc etc, to make any reasonable person happy.

This morning I stopped at the Columbia River Exhibition of History, Science, and Technology museum here in Richland, where for $2.50 a retired Hanford worker gave me a two-hour tour of interpretive displays and models of the Hanford Works. I quizzed a guy at work on interesting places to go see around here, and he was full of ideas. I ended up just driving up around Hanford Site to a place where there's a view of the reactors.

I drove out to the farming town of Othello just to have a destination... it started to snow, which at first was novel, but then very quickly I realized that there was a strong possibility that I'd get stuck out in the middle of nowhere. Was quite relieved to get to a larger highway which is plowed. I stopped in at LIGO to take care of a few things... the two people there were gearing up to be snowed in.

Back in Richland now... I have a flight out of here at 9:00 tomorrow.. to Salt Lake City to Chicago to Rochester. Rochester! Yay! Looking forward to seeing people I know again.

After a reasonably sumptuous breakfast at the hotel, Adrian and I drove out to the LIGO site. It's really closer than I had expected. The city of Richland, such as it is, peters out and quickly we're out on a straight stretch of 2-lane highway seemingly in the middle of nowhere. Steam is billowing up from somewhere a few miles off the road -- a reactor still running? The snow-covered desert is really quite pretty; the snowy hills blend right into the sky. At a bend in the Yakima river, we turn right onto the road that doesn't exist. A few miles later we're at the LIGO site, with a big sign out front with the LIGO logo.

There's a small collection of buildings at the "corner station" (the elbow of the "L"). One where we check in, where there are some offices too. Another has a tall roll-up door; this is some kind of assembly area. And the third building in the corner station itself. Inside everything is clean and well organised; it seems new.

The first half of the day was mostly the formalities (signing papers, reading safety information) and practicalities (getting keys, being shown around) of being introduced to LIGO, and the dozen-or-so people working on site. The Control Room is like Mission Control, with five projector displays on the wall (only the three forward ones currently in use), showing the current Interferometer status. Overhead, two large digital clocks display the time, one in UTC and one in local time. On two rows of desks there are numerous computer terminals, each with two large flatscreen displays. Almost all SUN Sparc machines. Closed-circuit TV displays show the area surrounding the LIGO center, and others show the LASER spots inside the interferometer. Most of the stations have the usual roller-chairs, but a few large inflated "fitness balls" are available too, lightening the atmosphere a bit.

At lunchtime we attended a meeting of the collaboration. Like at Big Physics labs everywhere, the meeting is done by teleconference on a sort of speaker phone system with PowerPoint slides projected onto the wall. The other LIGO observatory, in Livingston, Louisiana, is on the phone. As are, I think, the collaborators at MIT and Caltech. The meeting is, as meetings tend to be, very boring. They are planning when to do some big upgrades, planning for the next several years.

Daniel took Adrian and I into the clean-room environment that houses the corner station to get us acquainted. There's an automatic boot scrubber before we enter the 'garb' room, and then we put on booties and LASER safety goggles. Earlier Adrian had drawn for me a schematic of the LIGO observatory. The input laser is only a few watts, but the arms of the interferometer form resonant cavities, which build up a laser power on the order of kilowatts. The laser is infrared. We have proximity cards to enter the cleanroom area. We enter one-by-one, and the computer notes that we're now each "inside". Entering an area without using the proximity cards activates interlocks and automatically shuts down the lasers. The safety system is to protect people from lasers, and the cleanroom system is to protect the ultra-high-vacuum system and the clean optics from the dirt that rides in on people.

Once inside the clean area, the vacuum system of the corner station is visible. The beamlines are housed in a vacuum pipe about 1m in diameter. The optics systems are housed in large metal cylinders along the beamline that look like giant tin cans welded into the system. Beside the beamline are the usual rackmount systems familiar to any physics installation: crates containing VMW-bus backplanes, with cards sprouting patch cables. Analog signals are routed into digitizers. The digital signals are eventually stored in a sort of distributed shared memory, which is reflected between machines using gigabit ethernet over optical fiber. The air is clean and it's easy to breathe.

Daniel showed us around the electronics and we set about our mission. For various reasons, there is an interest in looking at a signal at 37.5 kHz -- noise tends to cancel out at this frequency, so it might be a good place to hunt for a signal. To gain access to this signal, software digital signal processing is used to shift this signal down to baseband, so that it can be stored at a reasonable sample rate. Our task is to verify the frequency response of some "whitening" (equalizer) boards, connect signals from the interferometer output photodiodes to this board, make sure the data acquisition (DAQ) is working properly, record data from this frequency range, and then look at it. The DAQ system seems to run on VxWorks (I believe the same OS that ran on the Mars rover).

I really have to brush up on my signal processing... FIR and IIR filters, Butterworth filter design, complex baseband signal representations / quadrature, control systems design. LIGO is really all about control loops: The main operating mode of the interferometer is to keep the interferometer output locked at a "dark fringe" (no light output; total destructive interference) by using actuators to move the mirrors; and the feedback signals to these actuators becomes the output signal of the detector. There are several resonant cavities, most of which are kept resonant by active control systems.

In between times, Adrian explains all about how LIGO works, and what we will be doing. ("And tomorrow, Dark Energy," he says. I don't know what Dark Energy has to do with anything...)