1 November 2012

The nature of noise

When Mother Nature whispers, physicist Albert Migliori listens.

By This email address is being protected from spambots. You need JavaScript enabled to view it.

Noise created by everyday itemsPowerlines, radio waves and telephone signals all produce electrical noise which can interfere with sensitive scientific experiments. Although we can shield experiments from interference, one kind of noise will remain constant: thermal noise generated by moving atoms.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Do you know a scientist? Old or young, student or professor, most experimentalists have a beef with noise. It’s messing up their experiment! There’s too much of it! The people who designed their equipment weren’t taking it into account!

When scientists complain this way, they (mostly) aren’t talking about acoustic noise — the sound of babies crying, pans rattling, car doors slamming or crowds cheering. They’re talking about the whisper of Mother Nature at work, says MagLab physicist Albert Migliori. We’ll dig into what this whisper is, why it matters, and how, complaining aside, noise can be a scientist’s friend.

Q: Albert, what is noise?

A noise graph.This graph depicts a simulation of “Johnson noise” (also refered to as thermal noise). When you remove all other noise — made by people, machines and the earth — what’s left is thermal noise generated by moving atoms. The fluctuating white line on the graph represents the pressure you feel in your ears when you hear the noise.

Noise! What a wonderful subject! I can talk about noise the way some people talk about love. When scientists talk about noise, they’re talking about something very different than you might think. You have to realize that there are some things that are so perfectly controlled by Mother Nature that they become immensely valuable to science. Noise is one of them in the very best sense.

Q: But what IS it?

The kind of noise I’m talking about sounds to the ear like a hiss. The hiss doesn’t have any particular quality about it. If I listen to the hiss 10 minutes from now it sounds exactly the way it will sound five minutes, 20 minutes, or two seconds from now. The hiss of noise is a steady background that sounds like a waterfall or running water.

Interference, which is the stuff that men and women make — man-made and electrical noise — would come from the spark plugs firing in a car that’s started outside the lab, or the graduate student texting on their cell phone and hitting send, or the radio station on campus transmitting in the same frequency noise where we’re trying to measure. Those are common causes of interference; we also get them from lightning during storms.

Q: So noise is a hiss?

After we get rid of interference — all that man-made stuff and lightning and that sort of thing — what’s left is the purest kind of noise. Mother Nature’s thermal noise comes to us because we are not at Absolute Zero in temperature.

GLOSSARY

Noise – A random fluctuation in an electrical signal. Noise determines the limit of the smallest thing measureable in an experiment

Signal-to-noise ratio – The proportion of useful versus not-useful information in an experimental process.

Thermodynamics – The study of the relationship between heat and other forms of energy — and where you’ve got energy, you’ve got noise.

Absolute Zero – The lowest temperature theoretically possible, in which absolutely minimal heat energy would be left in an atom. (Or you can just call it minus 459.67 degrees F.) Generally, the less heat present in your experiment, the less noise there is.

Entropy – The number of states available to a system at a given temperature.

Frequency – How often an event occurs in a given period of time.

That noise is really, really special. It connects to the most fundamental arguments about the arrow of time and nature, about thermodynamics, which is a set of approximations of how heat and energy are connected. And even though they’re approximations, they’re very precise and that’s one of the few phenomena in physics that tells you which way time is marching. Noise, thermodynamics and entropy are almost the same things, to a physicist.

Q: Wait, so do you like noise or not?

For the scientists at the MagLab, noise is our enemy because it clouds our ability to see the results of an experiment. Yet it’s also our friend because it tells us when we can stop trying to make the experiment better.

Mother Nature, in almost all circumstances, puts a quantitative number on what noise has to be. Mother Nature tells us that you can’t make things “quieter” than a certain level. That level is determined by the temperature of your experimental environment, and nothing else. If you’re a scientist at the MagLab, this is a relief.

Say you have an experiment and you have a lot of noise and you need to get rid of it. Somebody’s yakking in the hallway, you can close the door, right? But scientists have radio transmitters and cell phones and all this stuff in our modern world, generating electrical noise. You keep shielding and shielding, but then when you get down to Mother Nature’s limit, you can stop. You don’t have to keep working on it anymore because you can’t make it any quieter.

Q: Why is having a lot of noise a problem for a scientist?

When you’re driving down the road making a cell phone call (which you shouldn’t be doing) with the windows open (which is even dumber), the wind noise makes it hard to hear the voice on the cell phone. Then, the wind noise goes through the cell phone microphone and makes it hard for the other person to hear you. That’s exactly what noise does in the laboratory. It introduces random signals that are not what we’re trying to measure. Minimizing those random signals makes it possible to see something more clearly.

Now, the problem at the MagLab is that many of these signals are really, really, really tiny. It’s not like at a rock concert where it’s impossible to drown out the concert speakers. At the lab, all the experiments are whispering to us. That is often the problem scientists face at the MagLab, and to hear that whisper we need to make the rest of the room absolutely quiet. That means we have to get rid of every possible source of noise that we can get rid of. Because Mother Nature has given us the answer to what the minimum is, we know exactly how far we should go.

Q: What does the scientist do to control/ understand/ mitigate noise?

The physicist does exactly the same thing that the guy who built your car radio does. You’ve got a radio antenna sitting on your car. Say you turn it to an FM station. Why is it that you can hear that station and not the police, air-traffic controllers, cell phone conversations, and other radio stations? How in the world does that work? How does that antenna only pick out one?

It turns out that the amount of noise energy at any given frequency is fixed, and it’s all the same out to very, very high frequencies. If I can make a radio transmitter located on campus put out more energy at a particular single frequency than Mother Nature’s noise, then I win. In fact, even at 100 miles from a transmitter on a 100-foot tower putting out 50,000 watts, I can beat Mother Nature’s noise at that one frequency. So now I make a radio receiver that only senses that one frequency and doesn’t pick up anything else by doing a little bit of electronic magic. The magic is straightforward and we’ve been doing it this way for 100 years. It’s just a little technical engineering trick that says I can’t see anything but this particular radio frequency.

So the scientist does the same thing. Once you’ve gotten rid of the noise you don’t really want, the noise that remains tells us a lot about physics systems. The properties of noise, and the way they come out of some physical process that we’re trying to understand at the MagLab, are really powerful, because they’re so tied up with the most fundamental systems of thermodynamics and entropy. Anything that modifies that thermal noise tells us tremendous amounts about the physics if we’re sharp enough to use it. Looking at the noise itself becomes a powerful experimental technique.


This story was originally published in Issue 9 of flux magazine, a discontinued publication of the National High Magnetic Field Laboratory.