Table of Contents

The Hall Effect Gaussmeter


Hall effect gaussmeters are used in research, product design, education, and materials inspection. A better understanding of magnetic fields and the Hall effect will help the operator get the most out of these instruments.

John Boettger, F.W. Bell

Photo 1. Hall effect devices, or Hall generators, are made by attaching four electrical contacts to a thin square or rectangular plate or film of GaAs or InAs. A ceramic substrate is often used for mechanical support, thermal stability, and wiring nodes. Alternatively, the devices can be wire bonded to a nonmagnetic lead frame and encapsulated in a dielectric material. In this photo, the Hall generator is shown with a pencil point for scale.
Magnetic fields are well defined in theory but do not behave so predictably in real life. Often the root cause for the failure of a design involving magnetic fields is the designer's inability to understand how the lines of force, or flux lines, are generated or affected by the surrounding environment. In the centimeter-gram-second (CGS) system of measurement, one flux line is called a maxwell (Mx). In the more commonly used SI (international system), the weber (Wb) is 108 lines. Thus the relationship:

  1 Wb = 108 Mx

Figure 1 depicts a simple bar magnet with two poles. Faraday viewed the pole faces as containing thousands of smaller unit poles and the space surrounding the magnet as filled with the same number of flux lines, one line connecting each north-south pair. Flux lines are generally viewed as exiting the north pole and returning to the south pole. The total number of flux lines passing perpendicularly through a given area is called flux density, B, or magnetic induction.
Figure 1. All sources of magnetism have at least two poles that are linked by invisible lines of force, called flux lines. Today's sophisticated magnets are made in a variety of shapes, some containing many pole pairs generating complex flux patterns.
In the CGS system a gauss (G) is one flux line passing through one square centimeter. In the SI system the tesla (T) is 10,000 lines per square centimeter. Thus the relationship:

  1 T = 10,000 G

The force within the magnet that produces the flux lines is the magnetic field strength, H, or magnetizing force. In the CGS system, one oersted (Oe) is produced when two like poles of two identical magnets, placed one centimeter apart, cause a repelling force of one dyne. The SI system assumes an infinitely long coil of wire (solenoid) wound with a number of turns per meter and carrying one amp of current. The magnetic field strength at the center of the solenoid is given in amps per meter, or A/m. The relationship between oersted and A/m is:

  1 A/m = 79.58 Oe

It must be understood that flux density and magnetic field strength are related but not equal. The intrinsic characteristics of the magnetic material must be considered. Only in free space (air) are flux density and field strength considered equal.

The Hall Effect
In the 19th century Edwin Hall attached a wire to each side of a rectangular piece of gold foil (see Figure 2), passed an electrical current through the length of the foil, and measured the voltage across the width of the foil.
Figure 2. In 1879 Edwin Hall, a 24-year-old graduate student at Johns Hopkins University, discovered the Hall effect by experimenting with a thin sheet of gold foil and a current source.
Hall discovered that the voltage, now called the Hall voltage, Vh, is directly proportional to the number of flux lines passing through the foil, the angle at which they pass through it, and the amount of current used. He also found that the polarity of the voltage reverses when the direction of the flux travel reverses. Because of the extremely low voltage produced by the foil, the Hall effect remained a laboratory curiosity until the development of certain semiconductor compounds such as gallium arsenide (GaAs) and indium arsenide (InAs). These materials produce relatively large Hall voltages and led to the wide-spread use of the Hall effect in science and industry.

A modern Hall effect device (see Photo 1), commonly called a Hall generator, consists of a thin square or rectangular plate or film of GaAs or InAs to which four electrical contacts are made (see Figure 3).
Figure 3. A "bulk" Hall generator is constructed by soldering a thin slice of a III-V compound such as InAs to conductors on a ceramic substrate. Thin film devices are made by depositing the vaporized compound directly on the ceramic circuit plate.
The plate or film is often affixed to a ceramic substrate that provides mechanical support, thermal stability, and wiring nodes. Other devices are wire bonded to a nonmagnetic lead frame and encapsulated in a dielectric material.

The output of a Hall device is greatest when the flux lines are perpendicular to the surface of the material. When the angle is held constant, and a constant current is provided through the material, the Hall voltage is directly proportional to flux density. Conversely, holding the flux density and current constant allows the device to respond to the angle of the flux lines. One particularly useful aspect of a Hall generator is its ability to sense the direction of flux travel, allowing it to detect both static (DC) and alternating (AC) fields.

The active area, the area of greatest magnetic sensitivity, is considered to be located in the center of the plate or film and is the largest circular area that can fit within the boundary of the connection points. Present manufacturing methods have produced active areas as small as 0.13 mm in dia. Some devices are only 0.25 mm thick, allowing their use in very tight spaces.

The ideal Hall generator produces zero voltage in the absence of a magnetic field, but actual devices are subject to variations in materials and construction. Most Hall generators therefore produce some initial output in a zero field. This signal, known as the Hall offset voltage, Vm, can be canceled with external analog circuitry, arithmetically canceled by a computer, or removed by abrasive or laser trimming. The offset voltage is affected by temperature and can change in either the positive or negative direction. Vm is usually specified as a maximum ±µV/ºC change.

The ideal Hall generator has a constant sensitivity over a range of flux density, but actual devices are seldom linear. Typical accuracy ranges from ±0.1% to ±2% of reading. The sensitivity is also temperature dependent and always decreases as temperature increases for both GaAs and InAs. Typical values range from ­0.04%/ºC to ­0.2%/ºC.

A Hall generator produces a positive voltage for flux lines traveling in one direction and a negative voltage in the opposite direction. Ideally, for equal fields of opposite polarity a Hall device will generate equal voltages of opposite polarity. In reality there is a phenomenon called a reversibility error that causes these voltages to be slightly different in magnitude. This is caused, in part, by inconsistencies in the material's composition and by the locations and sizes of the electrical connections to the edges of the Hall plate. The error is usually stated in terms of percent of reading and can be as high as 1%.

A Hall generator's accuracy when sensing high-frequency AC magnetic fields is primarily limited by the connections to the Hall material rather than by the material itself. Inductive wiring loops within a changing magnetic field generate significant voltages on their own, sometimes higher than the Hall voltage. Careful design is required to reduce this effect to a minimum. (The Hall generator was also discussed in "Understanding Hall Effect Devices," Bill Drafts, Sensors, September 1997.)

The Hall Effect Gaussmeter and Probe
A Hall generator may produce signals as low as 500 nV/G or as high as 200 µV/G. Hall effect gaussmeters (or teslameters) are designed to amplify and condition these low-level signals and provide a result that is calibrated in terms of gauss and/or tesla. These instruments range from small handheld meters to more sophisticated bench-type units (see Photo 2).
Photo 2. Hall effect gaussmeters from F.W. Bell are available in a variety of single-channel and multichannel configurations in both digital and analog format. Hundreds of probe standard and custom configurations are also offered.
The instruments generally provide multiple range settings that allow the user to measure flux density as low as 1 µG (0.1 nT) and as high as 1 MG (100 T). Some meters measure only DC fields while others are capable of both DC and AC measurements. Some instruments accept input from several Hall generators simultaneously, allowing for 2- and 3-axis vector measurements. In addition to front-panel readouts, some gaussmeters provide analog signals that can be used in control loops or for analysis of pulsed or alternating magnetic field waveforms. High-end gaussmeters usually offer some form of communications, typically RS-232 or IEEE-488 (GPIB), allowing the instrument to be used in automated data acquisition and control systems.

Often the Hall generator is mounted inside a protective tube, or stem, made of aluminum, fiberglass or other nonmagnetic material. The wires are connected internally to a flexible cable and the cable is terminated with a multipin connector. This assembly, known as a Hall probe, is generally available in two configurations (see Figure 4).
Figure 4. Hall probes vary in thickness, material, and measurement ranges. Cables can be up to 60 m long. Mulitaxis probes combine transverse and axial measurement in a single stem.
Transverse probes are usually thin and flat; axial probes are cylindrical. The primary difference is the axis in which flux lines are sensed (B in the figure). Transverse probes are often used to make measurements between two poles of a magnet such as those found in audio speakers, electric motors, or MRI machines. Axial probes can be used to measure the fields generated by coils or solenoids. Either type can be used where there are few physical constraints. Some probes contain several Hall generators arranged orthogonally to allow simultaneous measurements in different axes.

The Hall effect is generally considered as having a maximum resolution of 1 mG (100 nT). Below this level, electrical noise and thermal effects swamp the usable signal.
Figure 5. A Hall generator's output is related to the number of flux lines passing through it. Ferrous concentrators can boost the signal by directing more of the local flux lines through the device.
Some gaussmeters use heavy filtering, modulation techniques, and sophisticated averaging in an attempt to provide better resolution. The signal can be enhanced by placing the Hall sensor near one or two pieces of iron or other ferrous material. These pieces, called concentrators, bend the local flux pattern so that more lines pass through the sensor (see Figure 5). Because of the local flux distortion and the size of the concentrators, this type of probe is normally used to make volumetric measurements such as in geomagnetic surveys, electrical interference studies, or preflight package inspections.

Using a Gaussmeter
Gaussmeters are nearly as easy to use as voltmeters, but there are several sources of errors that can affect accuracy if the operator is not familiar with the Hall effect or magnetic fields.

"Zeroing" or "nulling" the Hall probe and meter is one of the most important steps toward obtaining accurate flux density measurements. As stated earlier, most Hall devices produce an offset signal in the absence of a magnetic field. Second, the internal circuitry of the meter itself is likely to produce a small offset signal even in the absence of an input signal. Finally, local flux from the Earth (~0.5 G) or nearby magnetic sources will affect the Hall sensor. The process of zeroing eliminates these errors.
Figure 6. A Hall generator's output is related to the angle at which flux lines pass through it. Maximum output is achieved when the lines are perpendicular to the sensor. At other angles, the output follows a cosine function.
The probe is frequently placed in an assembly called a zero flux chamber to shield the Hall device from all local flux. In other situations it may be desirable to zero the probe without the chamber so that all future readings are relative to the local flux condition.

Another common source of error is due to the angle of the Hall generator relative to the flux being measured. As shown in Figure 6, the highest output is generated when the flux lines are perpendicular to the Hall sensor. This is the way each Hall probe is calibrated and specified. It is often incorrectly assumed that the plane of the Hall generator is exactly the same as the axis of the probe's stem, but because of variations in material and manufacturing this alignment is not a certainty. The user should always peak the probe, a process in which the probe is rotated and tilted in several planes to obtain the highest possible output for a given field. At that point the probe should be fixed in place.

Figure 7. Flux lines form a closed-loop pattern between poles. As the distance from a pole increases, the flux density decreases.
Hall effect measurement of permanent magnets can lead to confusing results. Flux density decreases as the distance from the pole face increases (see Figure 7). The Hall generator will always be some finite distance from the pole face because there will always be material (the stem and air) between it and the magnet. Flux lines are seldom distributed evenly across the pole face of a magnet. Interior flaws such as cracks or bubbles, or an inconsistent mix of materials, can result in flux density variations. The Hall device will respond to this if it is much smaller than the face of the magnet (see Figure 8). Finally, problems can arise from ferrous materials in the area where the test is being conducted.
Figure 8. Often the flux pattern across the pole face of a magnet is not homogeneous. A Hall effect gaussmeter can be used to map these variations and help designers improve on the magnet's design.
A steel workbench can redirect the flux lines from a magnet and cause erroneous results. Temperature effects, linearity errors, and reversibility errors should be taken into consideration when making Hall effect measurements. Modern gaussmeters can compensate for these problems, but the user should always refer to the specifications and take advantage of additional performance data if the manufacturer offers them.

Many gaussmeter manufacturers also offer a variety of permanent reference magnets and reference coils that can be used to verify the basic operation of the equipment. Verifying overall accuracy often requires a huge investment in magnetic standards and specialized equipment, so certification and calibration are often left to the original manufacturer or a third-party calibration lab. Most manufacturers recommend a one-year calibration cycle.

Hall effect gaussmeters provide an economical and relatively easy way to measure flux density. They are used in research, product design, education, and materials inspection. A better understanding of magnetic fields and the Hall effect will lead to more effective use of these instruments.

Magnetism: An Historical Perspective

The first recorded observations of magnetism occurred in the district of Magnesia, Thessaly, around 600 BC. Naturally occurring stones found in this region had unusual properties. They were attracted to iron but not to most other materials. Two stones might either be attracted to or repel each other. An iron needle touched by a stone would itself behave like the stone. If the stone or needle were freely suspended it would always orient itself to the same point on Earth. Thus the compass was one of the first practical devices based on magnetism, guiding the traveler as did a guiding star, or lodestar. The lodestone (magnetite, Fe3O4) derived its name from this analogy (lode meant way in Middle English), and the term magnet evolved from the district's name, Magnesia.

In 1269 Frenchmen Peter Peregrinus and Pierre de Maricourt, using a compass and a spherical lodestone, determined the existence of invisible lines of force surrounding the sphere just as the meridian lines surround the Earth, converging at points at opposite ends of the sphere. Maricourt called these points the north pole and south pole and noted that the force was always strongest at these points. Subsequent research found that magnetic poles always occur in pairs–if a lodestone is broken into many pieces, each piece will have a new set of poles.

In 1600 William Gilbert performed the first great systematic study of magnetism. Some of his most important work focused on terrestrial magnetism, and demonstrated that the Earth itself is a large magnet. At that time magnets were primarily used to lift heavy iron objects. Gilbert developed better ways to produce strong magnets, but the only magnetizing forces available at that time were other lodestones and the Earth. Major breakthroughs came in 1820 when Oersted proved a relationship between magnetism and electricity, and in 1825 when Sturgeon invented the electromagnet. Magnetic fields could now be generated at will and at intensities much stronger than any permanent magnet. These discoveries led Faraday to develop his theories on electromagnetic induction, which led to the development of the transformer, the alternator, and the dynamo. These inventions retired the chemical battery as man's primary source of electric current and led to the development of today's electric lights, television, audio speakers, credit cards, electric motors, toys, disk drives, medical imaging, levitated rail systems, and other wonders.


John Boettger is Design Engineering Manager, F.W. Bell, Division of Bell Technologies Inc., 6120 Hanging Moss Rd., Orlando, FL 32807; 407-678-6900, x-237, fax 407-677-5765.

We Love Feedback

Sensors® and Sensors Expo® are registered trademarks of Advanstar Communications Inc.


Sensors Online Home | Sensors Expo | Contact Us

Sensors Online