Some years ago, I was stunned to read that hydrogen protons can actually be heard with a (relatively) simple audio amplifier. At that time I was reading about magnetometers – devices that measure magnetic fields, and came across the “Proton Precession Magnetometer.” These things work by using a big electromagnet to magnetize a bottle of water (or other hydrogen-rich fluid). When the magnetizing current is switched off, the hydrogen protons inside will wobble about (“precess”) as they re-orient with the Earth’s magnetic field. This produces an electromagnetic signal which can be picked up by a coil of wire around the bottle, amplified and heard. The frequency of this signal tells us very accurately, how strong the Earth’s magnetic field is.
It sounded easy! I had to build one and hear this signal for myself. Trouble was there was not much available on the internet to tell me the easiest way to go about it. And, there was a lot that seemed to imply you had to spend a lot of time and money, and even then you still might not get it to work.
Three attempts at making one, and several years later, I finally heard it.
And this is the signal I hear from it:
Actually, the signal to noise ratio is better than my “good” one. I obviously still have more work to do!
Its Educational Value
At this stage, I have yet to get a clean enough signal to measure its frequency, but just to hear the signal is quite valuable to me. as it gave me a totally different viewpoint on my past physics studies. For me, it’s no longer a bunch of theory. It’s suddenly jumped out of those stuffy textbooks and into the real world, so to speak.
I dare say there are some science teachers who would like to be able to demonstrate such a thing to their students, as it has enormous potential for getting them interested in physics, and possibly even started out on an exciting career path. So I’m going to write about my experiences with this and how I finally managed to hear it. I won’t say it’s an easy thing to achieve, but if one is just trying to hear the signal, I think that given moderate skill with electronics and circuit construction, plus a bit of persistence, it’s very doable and not expensive. A full-fledged proton magnetometer though, is another story. It seems to require a high level of technical sophistication and is a definite challenge for the home experimenter attempting to develop his own from scratch.
While some designs use separate coils for magnetizing the core and receiving the signal, I got better results with a single coil. Apart from being simpler, it was also easier to get to it resonate at the frequency of interest, since the coil has quite a low DC resistance and therefore a higher “Q” than other experiments I did with thin wire. A spring-loaded toggle switch is used to either magnetize the core or listen for the precession signal. This gives an easy way to experiment with different time durations for magnetizing the core. The two capacitors, 0.47 and 0.039 uF are chosen to get a resonant peak at the frequency we expect for the local area. Those values, used with that coil, are right for Taiwan, where the precession frequency is around 1,900 Hz. They will probably need adjusting for your area. The JFET front end has a very high input impedance which lets the coil work as an effective narrow bandpass filter. This one falls off very fast outside the range 1850-1950 Hz. It’s a great way to reduce noise, and believe me, there’s a lot! It comes from power lines and electrical/electronic devices of all kinds. I used a 2n5485 JFET because I had some. One could use several others, such as the old MPF102, or 2N3819.
The BC547 amplifier stage runs at very low power, for minimum noise. Other transistors might be better, but one of my design criteria was that it should use commonly available, cheap parts. The two op amps are part of an LM324 quad op amp. Two of them remain unused. The thing was so ridiculously cheap… Both of them are configured as active bandpass filters for the range 1.5 – 2.5 KHz.
Notes on Getting a Precession Signal
The water in the core is distilled water. I bought a few different kinds of distilled water, including making my own. In the beginning, nothing seemed to work. Finally everything worked, even tap water (though it’s best not to depend on that). Bubbles in it don’t seem to matter. The plastic bottle in the core is best to be a little loose, so one can easily drop it out to make sure the signal is really a precession signal and not just feedback or some kind of resonance in the circuit. The coil must be placed East-West for the strongest signal.
For years, the absence of any precession signal remained a mystery. I made and re-made amplifiers, tried different fluids in the core, wound coils in every configuration I could think of. I could hear all kinds of things with the coil and amplifier -the weirdest of buzzes, clicks, whines, burbling noises, coming from almost anything electrical – fluorescent light fittings, my wall clock, computer screen, cell phone …. it seemed incredibly sensitive. But no precession signal.
I read somewhere that they won’t work inside a building, and yes, this is true. The first step on the road to success was when I started doing my experiments in a local park. When I finally made an amplifier with the right bandpass characteristics, I started hearing the signal. I never really imagined the signal would be so weak, but there it was, and then it was just a matter of finding ways to improve it.
There are two reasons why you have to get right away from buildings. the main one is the disturbance they make in the magnetic field. Even a tiny variation in the field through the core will kill it completely. If the field is not homogeneous, the individual protons won’t all be precessing in phase with each other and the resultant signal will be too small or non-existent. Steer clear of any big metal objects. You need to get ten meters or so away from a car and a few meters away from a bicycle. Roadside drains with reinforcing steel will kill it too if you’re near a road.
The second reason to get away from buildings is the electrical interference you’ll get from power lines. To make an amplifier that rejects noise of this type is quite difficult. It seems to be quite wide spectrum noise, and the simplest solution is to just get away from it. My favorite place for testing now is in some farmland about 100m past the end of the local power lines. The above recordings were done there. In any other local streets with power lines, it’s impossible to hear any precession signal at all.
To get this thing properly tuned, you need to firstly establish what frequency signal you should expect in your local area. You can find an approximation of your local magnetic field strength here:
Once that is established, convert that field strength (which is in nanotesla) to microtesla by dividing by 1000, then multiply by 42.576 to find the “Larmor frequency” in Hz.
E.g. from the chart, one can see Taiwan lies roughly on the 45000 nanotesla line, in other words, 45 microtesla. So,
45 x 42.576 = 1915.92 Hz.
Next we need a signal generator that will sweep across that frequency up to at least a few KHz higher. If you don’t have one you could make one with a 555. We also need to know its frequency to within a few Hz. An oscilloscope is a great tool to have here so you can see the resonant peak of the coil when you sweep over the frequency range. If you don’t have one, by ear will have to do.
I had mine set up like this, with a one volt peak-peak signal driving a 100 ohm resistor in a single loop nearby. The coil was quite close to the loop to begin with:
Moving the coil one meter away from the loop, with the coil tuned, I can just barely hear the signal. If you can hear it under those conditions, you should be able to hear a precession signal (when you take it to a good location.)
As time goes on, I intend to continue experimenting to determine the effects of various changes, such as the dimensions of the coil and water bottle (given in the circuit diagram). What I’m using is small compared to others I’ve seen described. Hopefully I can soon get a strong enough signal to reliably measure its frequency.