See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/238742630

Construction of a Proton Magnetometer

Article · January 2008

CITATIONS READS
2 9,831

2 authors:

Suranga Ruhunusiri Malagalage Kithsiri Jayananda

University of Iowa University of Colombo
45 PUBLICATIONS   705 CITATIONS    91 PUBLICATIONS   1,752 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Computer Vision View project

Development of novel ultrasound techniques for early detection of CKDu View project

All content following this page was uploaded by Suranga Ruhunusiri on 25 December 2013.

The user has requested enhancement of the downloaded file.

Proceedings of the Technical Sessions, 24 (2008) 78-85 78
Institute of Physics – Sri Lanka

Construction of a Proton Magnetometer

W. D. S. Ruhunusiri and M. K. Jayananda

Department of Physics, University of Colombo, Colombo 3

ABSTRACT

Proton precession magnetometer is a scalar magnetometer which is used in the total component
measurement of the magnetic field. This paper describes an attempt to develop a proton precession
magnetometer for the measurement of diurnal variation of the earth magnetic field in Sri Lanka. The
magnetometer constructed consists of a sensor in dual solenoid configuration, containing water as
the proton rich material, and an amplifier chain which was designed to suite the magnetic field
measurements in Sri Lanka and a polarizing and energy dissipation handling circuit controlled by a
Microchip PIC16F877A microcontroller. The proton precession signal which is in the audio range
was digitized using a sound card of a personal computer and the frequency of the proton
precession signal was determined using fast Fourier transform. 50 hours observations of the diurnal
variations of the geomagnetic field, carried out in a location at Wattala, Sri Lanka was used to asses
the functionality of the system.

1. INTRODUCTION

Proton precession magnetometer is a scalar magnetometer which is used for the

measurement of the total magnetic field of the earth. The measurements of geomagnetic
field is important for mineral and petroleum exploration, geological mapping, search for
buried or sunken objects, magnetic field mapping, geophysical research, magnetic
observatory use, measurement of magnetic properties of rocks or ferromagnetic objects,
archaeological prospecting, conductivity mapping, gradiometer surveying, and magnetic
modeling[1].
In particular, due to the fact that the geomagnetic equator lies over Sri Lanka, geomagnetic
measurements carried out in Sri Lanka are very important for geomagnetic and ionospheric
research [2,3]. However, only a few measurements carried out in this region have been
reported [2,3,4]. The proton magnetometer described in this paper was developed with a
view to fill this gap through the initiation of a geomagnetic research program.
The proton precession magnetometer is based on the precession of protons in a magnetic
field [5]. The magnetic dipoles of protons (hydrogen nuclei) contained in a sample of water
or a hydrocarbon are temporarily aligned or polarized by application of a magnetic field
produced by a current in a coil of wire. When the current is suddenly removed, the spin of
the protons causes them to precess about the direction of the earth magnetic field. This
precession of protons causes a small signal in the same coil used to polarize them. The
frequency of the signal is proportional to the total magnetic field intensity.

Construction of a Proton Magnetometer

Proceedings of the Technical Sessions, 24 (2008) 78-85 79
Institute of Physics – Sri Lanka

2. MAGNETOMETER CONSTRUCTION
Figure 1 shows the basic arrangement of the proton precession magnetometer constructed.
The main functional blocks were the sensor, polarizing and energy dissipation handling
circuitry and the amplifier chain. The frequency of the detected signal was determined by
Fourier analyzing the signal, after digitizing it using the sound card of a computer.

Figure 1: Functional blocks of the proton magnetometer

2.1 Sensor
The heart of the proton precession magnetometer is the sensor which consists of a coil of
copper wire wound around a container filled with a proton rich material such as water.
From the various configurations available for the sensors [6], dual solenoid configuration
was selected due to its property of cancellation of noise generated outside the coil.
In this configuration, two identical single coil sensors are placed side by side for sensing
the same magnetic field. The coils are connected in series so that the signals generated
inside them are added together. In this arrangement, any noise induced in the two coils due
to external sources will be canceled.
Two plastic bottles having a diameter of 5.8 cm and a coil woundable section of 7.8 cm
were used as the containers (figure 2). Plastic and glass are suitable substances for the
containers as they being non magnetic do not alter the local geomagnetic field. While water,
Benzene, Kerosene are suitable as the proton rich samples, double distilled water was used
due to its low cost and ease of use. The other reason for the use of water was its absence of
reaction with plastic.
The induced rms signal voltage from a practical sensor is in the order of micro volts. To add
to this problem is the Johnson noise generated due to the finite resistance of the sensor.
Hence, it is important that the sensor has good signal to noise ratio for a given bandwidth 
of the band pass filter. For a dual solenoid sensor which is constructed with identical
containers of radius rc length b wound with n turns of wire of resistivity  in ls number of
layers the induced rms proton precession signal Vrms is [7],

Construction of a Proton Magnetometer

 Proceedings of the Technical Sessions, 24 (2008) 78-85 80
Institute of Physics – Sri Lanka

 L  0 m nErc2 rw2 sin 

Vrms   sin gle (1)
2 2b ( rc  rw l s )
where  is the nuclear paramagnetic susceptibility of the proton rich material used,  0 is the
magnetic permeability of free space, E is the voltage of the polarizing supply used,  L is the
Larmor precession angular frequency of the proton precession signal and  is the angle
between the local geomagnetic field and the axis of the sensor.  m (filing factor) is a factor
introduced to account for the inequality of the polarizing field through out the entire cross
section of the sensor.
 m  exp(1.4rc / b) (2)

We have introduced an additional factor  single to account for the more fringing of the
fields at the four ends of the dual solenoid sensor when compared with its singles solenoid
counterpart.
 sin gle  exp(1.4rc / 2b) (3)

The signal to noise ratio (SNR) of the signal is given by,

 L  0 m n Erc2 rw3 sin 

SNR   sin gle (4)
2b[(rc  rw l s )  ]3 / 2 19.2kT

where k is the Boltzman constant and T is the temperature of the sensor fluid.

7.8cm

Figure 2: Partially completed sensor at right and completed sensor at left.

In the sensor design phase, the required gauge of the copper wire, number of turns and the
number of layers were determined with the help of the above equations in order to obtain an
rms voltage and SNR of close to 1 V and 100 respectively. In the design phase it was also
ensured that the resistance of the sensor was in the vicinity of 10 ohms because a high
resistance quickly dissipates the signal via joule heating. The inductance was kept at less
than 50 mH as higher inductance means that it takes a longer amount of time to turn off the
polarizing field, which leads to greater degradation of the signal. The specifications of the
dual solenoid sensor is shown in Table 1.
2.2 Polarizing and Energy dissipation handling circuit

Construction of a Proton Magnetometer

Proceedings of the Technical Sessions, 24 (2008) 78-85 81
Institute of Physics – Sri Lanka

The magnetic dipole moments of the protons cause the protons to precess in the magnetic
field inducing a time varying field in the coils. However, precession of all the protons in the
proton rich medium is not in phase. So for this to occur, first the proton rich medium is
subjected to a strong magnetic field by connecting the coil to a power source. This is
referred to as polarization. A 12 V battery was used as the polarizing supply for the
magnetometer. After waiting for the alignment of dipoles to take place which is governed
by the spin lattices relaxation time, the sensor is disconnected from the power supply. At
this point, the signal is picked up by the sensor and it is connected to the amplifier chain to
amplify the signal.

Table 1: Sensor Specifications

Parameter Value
Wire gauge SWG 23
Coil wound section length 7.8 cm for each container
Diameter of bottle 5.8 cm
Number of Turns Container 1 – 657 turns / Container 2 – 659 turns
Number of Layers 7 layers each
DC Resistance Measured - 16.9 
Inductance Measured - 31.6 mH
Polarizing Current (for 12 V supply) 0.71 A
Maximum current allowed 0.94 A
Filling Factor 0.7562
rms Signal Amplitude Estimated - 0.54 V
SNR (for a bandwidth of 250 Hz) Estimated – 61 dB
Resonant Capacitor 272 nF
(for a precession frequency of 1715 Hz)

Typically water has a spin lattice relaxation time between 2 to 3 seconds. So polarization of
water for 5 seconds yields a magnetization between 81% and 92%. Longer polarization to
achieve 100% polarization is avoided as longer polarization means the data gathering
interval becomes longer, resulting in reducing the number of data points gathered. After the
polarization, the magnetization of the protons decay in an exponential manner and hence
the signal disappears after few seconds. Hence the proton rich material has to be subjected
to polarization again. The signal pickup duration was selected as 5 seconds to reduce the
strain imposed on the relay in switching back and forth. So this arrangement yields a data
point every ten seconds. The switching based on these timing requirements were handled
by Microchip PIC16F877A microcontroller based circuit shown in figure 3.
Sensor is connected to SENSOR CON1 and SENSOR CON2 inputs. The RELAY CMD and
POLARIZE CMD are the commands applied to the GATE of T1 and T2 HEXFET s
respectively. Once The RELAY CMD and POLARIZE CMD are made high the sensor is
connected to the polarizing supply via the relay. After 5 seconds of polarization,
POLARIZE CMD is made low while keeping the RELAY CMD high. This forces the energy
stored due to sensor inductance to dissipate and this requires 2 ms. After waiting for the
specified amount of time RELAY CMD is made low and the sensor is connected to

Construction of a Proton Magnetometer

Proceedings of the Technical Sessions, 24 (2008) 78-85 82
Institute of Physics – Sri Lanka

amplifier. This allows the signal to be amplified and processed to extract the magnetic field
information. After waiting in signal pickup phase for 5 seconds the whole cycle is repeated.

Figure 3: PIC16F877A based polarization and energy dissipation handling circuitry

2.3 Sensor Tuning and Amplification

Prior to connecting the signal to the amplifier, the signal is allowed to resonate hence
slightly increasing its gain prior to inputting it to the amplifier chain. According to the
predictions based on the International Geomagnetic Reference Field (IGRF) which is a is a
mathematical description of the Earth's main magnetic field developed by the International
Association of Geomagnetism and Aeronomy [8], the nominal geomagnetic field total
component in Colombo region is around 40300 nT which corresponds to a Larmor
precession frequency of 1715 Hz. With the expected Larmor precession frequency and the
measured inductance, equation 5 was used in the calculation of the value of the resonant
capacitor required for a particular sensor. Resonating the sensor has an additional advantage
of providing initial band pass filtering. The required capacitance was obtained by
connecting several Mylar capacitors in parallel configuration.
1
C r= 2 2 (5)
4π f L L

The amplifier constructed consists of three separate but complementary sections - pre
amplifier, band pass filter based on multiple feedback band pass filter and a final gain
amplifier, as shown in figure 4. The amplifier chain was based on a low noise amplifier
OP77 with a noise voltage rating of 10 nV/(Hz)1/2 at 1 kHz. Metal film resistors were
primarily used as resistors throughout the amplifier chain to reduce the addition of Johnson
noise. The center frequency of the amplifier chain was 1778.3 Hz with a bandwidth of 173
Hz. The largest geomagnetic field variations that can be expected is during magnetic storms
and they can be as large as 1000 nT (which correspond to 43 Hz). Therefore a band pass

Construction of a Proton Magnetometer

Proceedings of the Technical Sessions, 24 (2008) 78-85 83
Institute of Physics – Sri Lanka

filter having a center frequency of 1715 Hz and a bandwidth of 86 Hz may be the best
choice. But the geomagnetic field predicted by the IGRF model is merely a guidance and
not an absolute reality as various external sources can further modify the local geomagnetic
field. Hence a much larger bandwidth was used for the band pass filter to increase the
chances of detecting the proton precession signal.

Figure 4: Amplifier chain used in the Proton precession magnetometer

2.4 Signal Analysis

Because the proton precession signal is in the audio range, a sound card of a laptop
computer could be used for digitization and the resulting data were analyzed using a
software package called Spectrum Lab [9]. The sensor was placed so that it is perpendicular
to the local geomagnetic field in order to obtain a signal of maximum rms magnitude
(according to equation 1). Sri Lanka has a geomagnetic field declination of – 2.9 degrees
which implies that the field vector is almost parallel to the earth surface hence the sensor
was placed so that its axis is perpendicular to ground.
The preliminary investigations conducted at a location in Nugegoda revealed a FFT peak
corresponding to 1780 Hz (geomagnetic field of 41823 nT). The authenticity of the signal
A
was verified by bringing a screwdriver which has the effect of increasing the gradient along
the sensor which result in the disappearance of the signal and the reduction of the signal
amplitude when the sensor axis is oriented parallel to ground.
E
Permanent noise peaks picked up by the sensor were visible as stripes in the waterfall
display. The proton precession signal appeared as a dot in the waterfall display due to its
B
C D
Construction of a Proton Magnetometer

FFT peak of proton precession signal

Proceedings of the Technical Sessions, 24 (2008) 78-85 84
Institute of Physics – Sri Lanka

decaying due to spin lattice relaxation and this further verifies that the FFT peak observed
is indeed the proton precession signal (figure 5). The duration of the proton precession
signal was found to be between 0.7 and 0.8 seconds with the help of DSP.

FFT peak of proton precession signal

Proton precession signal as a dot in

waterfall display
Figure
5: FFT peak associated with proton precession signal.

The geomagnetic field value was determined via the Larmor relation,
B0 = 23.496241  f (6)
where B0 is the geomagnetic field in nT and f is the frequency of the proton precession
signal in Hz. The geomagnetic field was calculated by considering frequency of the proton
precession signal as the frequency associated with the FFT peak. The accuracy of the result
is primarily governed by the resolution which is the ratio of sample rate to the FFT length
used. Data was obtained at a location in Wattala using a spectrum setting of 11,025 sample
rate and a FFT length of 16,384 which proved to be the optimum setting that could be used
in Wattala to ensure that the peak detected was always the FFT peak associated with the
proton precession signal and not a permanent external noise peak picked up by the sensor.
This setting gives an uncertainty of  4 nT for the calculated field values.
The observed geomagnetic field variation observed in Wattala (figure 6) lies between
40,040 nT and 40,220 nT which is consistent with the predicted IGRF model [7]. Based on
hourly means, the peaks for day 1 and day 2 were observed at + 89 nT and + 53 nT
respectively from the overall mean of 40,096 nT for 50 hours.

3. CONCLUSION

The results described above indicate that the developed proton magnetometer is capable of
measuring the earth magnetic field with sufficient precision for the study of diurnal
variations. However, for studies of geomagnetic micropulsations, the precision needs to be
improved further.

The main difficulty that had be faced in the development of this instrument was the low
signal to noise ratio of the precession signal. The approach used in overcoming this

Construction of a Proton Magnetometer

 Proceedings of the Technical Sessions, 24 (2008) 78-85 85
Institute of Physics – Sri Lanka

difficulty was to digitize the signal and employ digital signal processing techniques, which
required the use of a laptop computer. In order to make a more portable instrument, the
digital signal processing component will have to be developed in an embedded computer.

40096 nT

Figure 6: The diurnal variations of the geomagnetic field from 05.30 hours (GMT) on
17.05.2007 to 19.30 hours (GMT) on 19.05.2007 at Wattala. (raw data)

REFERENCES
1. S. Breiner, Application Manual for Portable Magnetometers, Geometrics, San Jose
California, USA (1999).
2. R G Rastogi, Talichi Kitamura and Kentarou Kitamura, First observations of
geomagnetic field variations within equatorial electrojet belt in Sri Lanka, Indian
Journal of Radio & Space Physics, Vol. 33,(2004), pp.83-87.
3. R. G. Rastogi, T. Kitamura, K. Kitamura, Geomagnetic field variations at the equatorial
electrojet station in Sri Lanka, Peredinia, Annales Geophysicae (2004) 22: 2729 – 2739.
4. M. L. T Kannangara and P. C. B Fernando, Nighttime Equatorial Pi 2 Micropulsations,
Journal of Geophysical Research, Space Physics, Vol 74. No 3 (February 1969).
5. L. Huggard, Proton Magnetometer, Practical Electronics, Wimbourne Publishing,
U.K.,(Oct 1970).
6. W. Bayot, Practical guidelines for building a magnetometer by hobbyists,
http://perso.infonie.be/j.g.delannoy/BAT/PPMGuidelines.htm.
7. J.A Kohler, Proton Precesssion Magnetometer, Comox Canada (November 2004)
8. International Association of Geomagnetism and Aeronomy,
http://www.iugg.org/IAGA/iaga_new_home.htm.
9. http://freenet-homepage.de/dl4yhf/spectra1.html.

Construction of a Proton Magnetometer

View publication stats