About twenty years ago a Summer student I had a few years earlier called to talk about digital cameras. She was waiting for the cost of "good enough" sensors to drop, but was excited by the prospect of eliminating the film costs. Soon we were talking about the total cost of taking a photograph. There was the cost of a trip if you wanted one from the North rim of the Grand Canyon and a photo of the Earth from the Moon would cost billions - less if a robot took it. There was an interesting optimization question: how far up could you go to take a photograph for a fixed amount of money? One of us suggested five hundred dollars. An airline ticket could get you to nearly 40,000 feet with some planning. But a weather balloon? Now we had something to think about! An hour of dreaming and scheming quickly went by. It was a fun little gedanken project to play around with. I moved on to other things the next day and mostly forgot about it. She didn't.
A month later, the phone rang - something about scoring a dozen "Barbie's First Cameras" for ten dollars. "That's four hundred and ninety dollars to go...you in?" The cameras were cardboard disposables with with a plastic lens and shutter assembly and about a dozen exposures worth of color Kodak film. They were pink. Of course I was in.
I found a damaged Trimble GPS with a broken screen for fifteen dollars. The electronics worked so it was only a matter of getting the signal out and sending it to the ground. She was responsible for programming an ancient Z80 microprocessor and working out a mechanism to push the shutters. I repurposed an old amateur radio. We modified a small styrofoam box with a battery and resistors to keep everything warm. Then it was finding a weather balloon and helium.
This leaves out dozens of details. Many were ugly, but good enough rules. This would be easy these days with electronics available to hobbyists. You could do a much better job for under a hundred bucks without the balloon, or just send up an old smartphone.
Three of the cameras worked. This is the best photo from 26,900 meters above Ohio.
The goal of engineering is to find a good, hopefully great, optimization to a problem with several competing requirements. Usually it's impossible to optimize everything. A car might have cost, performance, fuel economy, safety, time to market, and appearance as desirables. There's no way to maximize everything. Usually the problem is constrained - a supercar throws cost and fuel economy out the window and turns out to be easy to execute. Mass market cars are much more difficult. Great engineers excel at this type of optimization. I don't have the necessary talent and experience. Fortunately I live in a different problem space.
Experimental science is generally underfunded. Not only is it difficult to justify buying some gizmo, but sometimes the gizmo you really need doesn't exist. You need to modify something you already have or can borrow or just build something. The result is often crude, but it works. Many physics departments make sure you can qualify for a key to the metal and electronics shops. These days knowledge of 3d printing is almost assumed. More specialized skills and rules of thumb come with what amounts to the apprenticeship otherwise known as the PhD program.
I don't have an exhaustive knowledge of homebrew equipment, but you see it almost everywhere. Here are a few I've used or have built
° We had a particle detector with an internal atmosphere of argon, krypton and xenon. A cheap way to measure their relative concentrations is to measure the speed of sound in the gas. In the 70s Polaroid had a camera with an ultrasonic rangefinder. It sent out a short ultrasonic pulse and timed the return of the reflection to adjust the focus. Damaged cameras were cheap and we had existing electronics that did a much better timing job. There are probably a dozen other ways to do this, but it worked. I later used one of these transducers to listen to bat calls. It led to a series of bat detectors I built as a hobby. Again - not elegant, but workable and I learned enough to ask informed questions about bats.
° If you need a high voltage, low current power supply, the flyback transformer in an old CRT color TV gives you about 25,000 volts and is free if you can find one in a junk pile.
° Many experiments need to minimize the effect of the Earth's magnetic field. Usually a set of three Helmholtz coils - three pairs of circular wire loops current runs through. Each pair generates a uniform magnetic field. Three of these pairs, each at right angles to each other, allow you to create a field that counteracts the Earth's field. I once used Hula Hoops as forms to wind the wire.
° Immersion cookers (sous vide) in a styrofoam lined box are a good way to keep a liquid bath at a stable temperature above room temperature. Much cheaper than commercial heated water baths.
° Some biology labs I've visited use household refrigerators with plastic baffles and better thermostats saving thousands per over lab grade refrigerators.
° The performance of a small anti-vibration table can be improved by "floating" it on powerful speaker magnets with coils connected to an audio amplifier and simple feedback circuit.
° Home photo scanners (Canon, Epson, whatever) can be hacked to build the guts of a table that can position something lightweight in two dimensions to better than a tenth of a millimeter. An "x-y" table. For the price accuracy and repeatability of these scanners are remarkable.
° At Bell Labs in Murry Hill, NJ a prototype x-ray lithography tool created a x-ray beam by shining a beam of electrons on an iron target. The target? - a VHS tape driven by the tape mechanism taken out of a VCR. The iron particles embedded in the tape were perfect for the job. After about ninety minutes a fresh blank tape would go in.
° Back to my student days at Brookhaven National Laboratory. One experimental area produced a clean beam of neutrinos. A beam of protons with a target producing a lot of particles including neutrinos. Neutrinos rarely interact with matter - a beam of them would lose about half its intensity after passing through through a wall of lead a light year thick. The problem is some of the other particles will travel several kilometers through the air. The trick is to put a lot of mass between where they're produced and your experiment. Somehow BNL managed to convince the government they should store about 50,000 tons of armor plates from WWII ships in a line at the lab ... you know, just incase there's a strategic need for iron in the future. It's covered with dirt and has nice wild flowers growing on it.
° Commercial camera lenses can be amazing. The University of Toronto's Dragonfly array uses two clusters each comprised of 24 400mm F2.8 telephoto lenses to make the equivalent of a 1 meter f 0.4 refractor. The later would probably cost more than ten times as much - if it was possible to even make it. It does things other instruments can't and has made important contributions to astrophysics.
I'll stop here. Experimentalists often have to be cheap. Sometimes it's invention, sometimes it's just getting something working. It's definitely not engineering - a good engineer would throw up their hands.
Comments