Equipment failure

By Marcus Wilson 13/08/2012

In the last couple of weeks, I’ve been working with one of our technicians tracking down what has been going wrong with one of the experiments we get our third year physics students to do. It’s on Brownian Motion. Specifically, analyze the movement of small particles suspended in water by scattering of laser light. By studying the way in which the scattered light varies in intensity with time, we can work out the size of the particles in the suspension.

So says the theory. However, in practice the pattern of scattered light is nothing like what we’d expect in this situation. There was clearly something going wrong, but working out what hasn’t been straightforward.

In the end, we just went through piece by piece through all the equipment, and the interfaces between the equipment, checking each was doing what it should have been. In the end Stewart worked it out – we had a dodgy oscilloscope. It’s rather easy to trust your instrumentation, especially that you’ve paid a lot of money for, but it is worth remembering that sometimes it breaks, and, when it breaks, it might not do so in a manner that is obvious. A piece of equipment that spits out the dummy and refuses to do anything is rather less dangerous than one that, on the face of it, is doing its job, but actually is getting it wrong. In this case the consequences of the fault are hardly serious – we’ve just had an experiment that was clearly giving puzzling and unbelievable results. In fact, for the last couple of years, I haven’t had the students even attempt it, because I’ve known something’s been amiss with it. However, that’s not always the case.

There are similarities I think with those faster-than-light neutrinos that hit the headlines last year. It was a crazy result – hence the attention – but on the face of it the experimental results appeared to be real. But, very careful checking of the apparatus highlighted a couple of glitches with the equipment. It wasn’t doing exactly what it was supposed to be doing. The problem was small, but it was big enough to produce a sensational result.

Fortunately, science has ways of correcting itself, and in due course the problem was tracked down by some careful investigation. It’s interesting that this is a skill we often overlook in teaching our students. In an effort to illustrate the theory, we present them with experiments that actually work (or at least try to). We never (certainly not here) deliberately give students dodgy equipment and then teach them how to find out what’s going wrong. Given that it is a skill that any experimentalist needs to have, it should be one we teach. Something to try in the future with another class of guinea pigs.

Site Meter