Digital Diary of Ben Schwartz

Today I discovered that most of my results from yesterday were irreproducible, and that my basic sanity checks all fail. My oscilloscope continues to show voltage even when the probe isn’t touching anything, through capacitive coupling. If I short out the output of the amplifier, my probe still indicates 7000 Volts across the short circuit. Then the amplifier suddenly died, and I gave up and went home.

Today was the first day of classes. It is really difficult to figure out which classes to take, especially given that they are scattered cross three widely separated campuses, one of which contains my lab, which I’m supposed to be spending time in.

Today I started trying to figure out how to wire up my marginal homemade capacitor for high voltage. The setup was basically a jumble of bare wires and alligator clips with some copper plates. I wanted to measure the voltage across the capacitor. My oscilloscope tops out around 20 Volts, so I rigged up a 1000:1 voltage divider which I placed across the capacitor. I ran my signal generator at 5 MHz and connected my capacitor directly to the output of our usual military-grade amplifier.

I turned the signal generator voltage up until the oscilloscope showed 1 Volt. That means that there was 1 KV across the capacitor at 5 MHz. I could have gone maybe twice as high before hitting the amplifier limits before I chickened out. Then I put in an inductor with a resonance at 5 MHz. I hit 1 KV with over a factor of 10 in headroom. I could probably have gone to 10 KV at 5 MHz. I have no idea what that would do. It might ionize the air around it and start glowing. The FCC might get mad. Nearby objects might begin to heat inductively.

This evening I initiated a dinner for the Biophysics first-years on Spring Semester’s Eve. I wanted to know what classes people are thinking about taking, and I think I’ve found some neat ones. We’ll see.

I got my sonar program working, kinda. It gave wildly wrong answers, but at least it gave an answer. I suspect that I could make it actually work without a lot more difficulty, given two computers over which I have total control, but that’s a job for another time.

To say that details are unimportant would be an understatement. Today I attempted to make phantoms for my experiment. It’s hard to explain what a phantom is. In the world of imaging, you could say that a phantom is something to take a picture of instead of a person. It’s usually a big block of something like Jell-O (in this case, polyacrylamide gel). I’m not really doing an imaging experiment, so the phantom is more like a target with material properties that very vaguely resemble animal tissue.

Anyway, the chemistry worked, but the physics didn’t. I was making the phantom in test tubes, planning to let it solidify with strategically placed ultrasound-absorbing rubber at the bottom of the test tubes. The rubber sinks in water, but surprisingly it floats in the Jell-O solution, so I have phantoms with rubber at the top instead of the bottom.

Grr.

Today at work, I began building my apparatus. Thanks to a combination of luck, loose requirements, and flexible thinking, I was able to cobble together a suitable piece of equipment out of spare parts with almost no actual construction. Only a tiny bit of soldering was required.

Tomorrow I have to learn how to make electrically conductive Jell-O.

I felt pretty unproductive at work today, but I got a few things done: notably, making an appointment to set my classes for next term. After work I got back to work on my program, with a brief interruption for dinner and some CD recording.

I should tell you what my new program does. It’s fairly simple, really: it measures the distance between two computers. It does this by making a sound with the speakers of one computer and listening with the microphone of the other. There’s a lot of sophistication, but the bottom line is, it measures the amount of time it takes for sound to get from one laptop to the other and multiplies by the speed of sound.

“Does” is a bit an overstatement. It might. I only have one computer that I can use to test it, and I need two, so I have no idea if it actually works. In fact, the chance that it works is really quite slim. Nonetheless, it’s complete, and debugging it shouldn’t be too hard.

You might well ask why anyone would want to use computers to measure distances, especially given that accuracy of this technique is likely to be fairly low, perhaps a 5% margin of error. Surely it would be better to use a tape measure and save several thousand dollars. You’d be right. But the new One Laptop Per Child project is going to create an unusual situation, where millions of (young) people own well-equipped computers but lack basic education supplies like rulers, notebooks, and chalk. Thus, I am building a software ruler. There might also be other advantages, like being able to work over distances of 20+ feet, which no ruler and only extra-long tape measures will do. But that’s only if it works.

There was some beautiful snow last night, and throughout the morning, but not enough to amount to a snowstorm.

Tonight was Bush’s State of the Union address; it seemed kind of like Geraldo, with all the celebrity guests.

I’ve finished the first stage of my design for an acoustic range-finder thingy. The details are not important, but at the moment, in theory, my software can measure the sonic impulse response of my room. It seems to be working, which is very exciting, and puts me in a very good mood.

That’s especially good because today was shaping up to be a bad day. I had an appointment with the Chorallaries for a very cool gig: to perform “Happy Birthday” in Norwegian on videotape on behalf of the MIT Alumni Association as part of a birthday gift to a Norwegian alum. I moved everything around in my schedule, arrived at the recording studio, and it was locked. I called everyone in the Chorallaries and no one picked up. I checked my e-mail on my phone and discovered that I had misunderstood a previous e-mail and we were supposed to have met an hour earlier, in a room on the other side of campus. So I went across the campus, but no one was there. I search the entire floor, then went back to the studio, only to find that the Chorallaries were there. They had been inside the room the whole time, with their phones off and the doors locked.

When I got home I started working on my project to try to reduce my frustration, but I ended up accidentally breaking the sound on my computer, which took me several hours to get working again. Then I didn’t have enough time to eat dinner before the Bad Taste writing session at 7, so I didn’t eat until afterward, at 9. I over-defrosted my frozen beef, and I had to throw out the partially-cooked bits, though I suppose I probably could have figured out a way to cook them in.

Now it is definitely sleep time. Hopefully I’ll get to dream about my tiny beautiful new program instead of the mess everything else has been.

I haven’t had much to do this weekend, so I’ve been puttering, which is fine, but when I found that I had reverse-engineered the kernel driver for a piece of hardware that I will almost certainly never own, I knew drastic action was needed. So I went for a run.

It’s 25 degrees outside, so I ran in (literally) head to ankle spandex, my wonderful new running shoes and the heavy nylon zipper pants from my high school swim team. My hands were awfully cold by the end, even in gloves, but it was worthwhile. There are few things in my world that are quite as beautiful as the Boston skyline at dusk from across the Charles, and sometimes you need to go for a run just to see it.

A few days ago, I was riding my bicycle down (snow-free) Mass. Ave. when I was passed by a late-80s Toyota Camry covered in snow. It had not snowed in Massachusetts in a month at least, and I don’t recall any weather reports of snow anywhere nearer than Colorado. So if you were driving through Cambridge covered in mysterious snow, I saw you.

Today it kind of snowed. It was more like slow rain, and it’ll probably be warm enough tomorrow to bicycle around, which is good because I have some bicycling to do. I found an article that looks like it might be helpful for my research, but it was published in 1981 and Harvard doesn’t have a copy of it. MIT does, in some storage archive off-campus, but I’ll probably have to make an inter-library loan request through Harvard, which will then have to be pulled from the archives at MIT.

I got my project today. Specifically, I had a meeting with my PI and our collaborator, who presented a very impressive outline of a cool new technique that I probably shouldn’t talk about too loudly. My job is to see if it can work, how it might work, and try to test it inside of 3 months.

To that end I spent today on a literature hunt, eventually taking out the famous Landau and Lifshitz but also a monograph entitled “Field Matter Interactions in Thermoelastic Solids.” It was published in 1975. It was in the Harvard Physics Research Library. I am the first person ever to take it out.