FW Bell Gaussmeter Probe Selection and Application Guide
Larry Purvis July 26, 2017
When selecting a gaussmeter and probe several things must be considered to determine the best gaussmeter
and probe models for the application.
1. The range of flux density to be measured.
2. Is a transverse or axial configuration required?
3. The type of signal to be measured; DC or AC?
4. If AC
a. What frequency or range of frequencies are of interest?
b. What type of waveform, sinusoidal, square, unidirectional pulse etc.?
c. Continuous signal or one‐shot / burst / pulse etc.?
5. What accuracy (absolute or relative) and precision or resolution is required?
6. Environmental conditions; temperature, humidity, vibration, handling stresses
7. Do you require analog output signals?
8. Are digital outputs or remote control capabilities such as RS‐232, IEEE‐488, USB or Ethernet
required?
Probe selection
The first thing to determine is the basic probe configuration. In many cases either a transverse or axial probe
may be used if there are no physical limitations which force you to use one or the other. If for example you are
just checking residual magnetism on a pipe or measuring the surface of an object either type will probably work
as long as you can orient the probe stem and cable to a position which allows a proper reading.
Transverse probes are required if you need to measure the flux density in a gap between objects or surfaces
such as in a cut core, between rotor and stator of a motor, or perhaps a loudspeaker magnet assembly.
An axial probe would be required if you need to measure inside of an object like a small diameter solenoid coil
or sputtering chamber.
See the following photos for examples:
OECO LLC, A Meggitt Company Tel: +1 (503) 659 5999
4607 SE International Way, Milwaukie OR 97222, USA Fax: +1 (503) 653 6310
www.oeco.com
OECO LLC
www.meggitt.com
Transverse probe required
Photo 1 -Core Gap Photo 2 -Between objects
Axial Probe Required
Photo 3 -Solenoid Coil
Either Axial or Transverse -External Surface or Free Space measurements
Photo 4 Photo 5
As can be seen from photos 4 and 5, the same measurement could be taken with either probe
configuration since there are no physical limitations. In both cases the Hall plate inside the probe is
parallel to the end face of the pipe. If there are no other limitations or reasons to choose a particular
configuration, the axial is usually a little more robust to handling due to the larger cylindrical probe
stem. This may be a concern in certain applications where there is a higher risk of handling damage.
Specifications and Accuracy
Most probes have a linearity specification which is expressed as a percentage of reading error over a
range of magnetic flux; for example 1% to ±30kG, or 0.5% to ±2T. This specification is for DC fields and
must be added to the accuracy specification for the gaussmeter on the range being used to determine
total measurement uncertainty. Some probes such as the ZOA 3‐axis types have a linearity range
specified to a flux density level which is less than the full scale reading of the gaussmeter (10kG for the
probe and 30kG for the gaussmeter). Most of these will still respond to higher field levels, but with
increasing error above the upper specified limit.
The exceptions to this are the Low Field probes. The MOS models utilize Fluxgate technology and their
internal circuitry will clip and not measure beyond the ±1.0G limit specified. The model MOX or MOW
Hall probes (“MagnaProbes”) with flux concentrators will respond above the 2.0G specification limit, but
will become non‐linear very fast between 2 and 3 gauss and may require degaussing to restore accuracy
if operated beyond ±4 gauss.
Accuracy for AC fields is not as good as for DC. This is because the signal is susceptible to more parasitic
effects as noted in the AC Considerations section below, and the fact that finding and constructing a
primary reference standard for AC is more difficult than for DC. Since applications can vary greatly, we
do not have hard specifications for AC. We attempt to compensate for the major factors to keep error
within a few percent, but each environment and a probe’s response to it is different. In most cases the
compensation will keep the error well below the ±3dB limits often used for AC bandwidth specifications
on measuring equipment.
Geometric Tolerances as a Source of Variation
In situations where the user is positioning the probe by using the outer surface of the stem as a
reference point and measuring on the surface of an object there may be differences observed between
probes of the same model. This is due to manufacturing tolerances which will place the Hall device at
slightly different positions within the probe. In most cases this is not of significance, but in others this
needs to be considered. For example the Hall plate in a transverse probe may not be exactly centered in
the stem and if the readings with one side against the surface are compared to those with the opposite
side of the probe on the surface there may be a difference in magnitude. The polarity will obviously be
opposite, but the magnitude could differ too. Even though this positioning error is only a few
thousandths of an inch, in some applications it is necessary to be aware of this and compensate
accordingly. Also the Hall plate may not be exactly parallel to the probe stem surface and relying on the
outer surface of the probe stem for positioning may result in a reading which is slightly lower than the
actual flux density due to this angularity error. Probes are calibrated in an air gap and rotated until the
peak reading is obtained.
Thermal Considerations
Temperature coefficients of the probe should be considered if operating at temperatures more than a
few degrees from the usual ambient of ~22°C and high accuracy is required. Temperature compensated
probes may be helpful in these instances and these corrections are performed automatically by the
gaussmeter. In other applications which are outside the normal compensation range (such as cryogenic
probe usage) manual correction will be needed. Temperature coefficient data for most probe models
may be obtained from their data sheets. If it is necessary to measure an object which is hotter or colder
than the specified operating temperature range for the probe, consider shielding the probe stem in
some type of non‐magnetic tube which has air or fluid flowing through it to keep the probe at a more
moderate temperature. This can also increase accuracy even if you are within the specified operating
temperature limits for the probe and do not have a temperature compensated probe.
Thin Film Probes
There are several models of thin film probes available for the 8000 series meters. These probes have 10
to 20 times higher impedance and therefore have a little bit higher noise floor than the normal bulk
types. Typical noise or short term drift for an 8000 series thin film probe is about 5mG. This is a bit of a
tradeoff to get the smaller sizes available with this type of probe. Thin Film probes are also available for
the 5100 and 6010 meters, but due to the lower display resolution of those meters, the additional noise
is not noticeable.
AC Considerations
In general, errors from the following parasitic effects will increase proportionally as frequency increases.
These apply to both continuous and discontinuous AC signals (pulses).
• Aluminum probe stems may have some error at frequencies above a few kilohertz due to eddy
currents developed in the stem. Some axial probes have a small slit in the tip to reduce this error
and are still relatively accurate up to 20kHz. The probes designated as “heavy duty” have model
numbers beginning with an “H” do not have this slit, are thicker aluminum and may exhibit more
eddy current error.
• There is always some finite loop area associated with the Hall plate itself, wiring in the stem and
junctions at the Hall element and probe cable. We attempt to keep such areas as small as
practicable, but they can’t be totally eliminated. These loops when exposed to an AC magnetic
field will generate inductive voltage signals which mix with the output of the Hall sensor. This
inductive voltage increases more‐or‐less linearly as the frequency increases. The probe body or
handle is usually the most susceptible area to external field influences and should be kept out of
the field if at all possible.
• The amount of the probe stem, probe body and cable which is exposed to an AC magnetic field
can affect the reading. We calibrate probes for frequency response in an attempt to increase
accuracy over a wide range of frequencies, but each application is different and we can’t
compensate for each one. We do use consistent methods which we believe to be typical of most
applications when calibrating for AC frequency response. Transverse probes are calibrated with
only about 0.5” of the probe tip exposed to the AC field. Axial probes are calibrated with 2” at
the tip of the stem inserted into the AC field.
• Depending on the angle of any inductive loop compared to the Hall plate the angular position
within an AC field may not give exactly the expected results. For example, if the probe was
positioned with the Hall device at a 60° angle to the field direction one would expect a reading
about 86.6% of the value when the Hall plate is at 90° to the field. Any inductive signal could mix
with the Hall signal and alter this value or produce signals which would be the vector summation
of the Hall and inductive signal.
Using a gaussmeter to measure low currents
For short term experiments a gaussmeter could be used with a gapped toroid core to form an open loop
current sensor. This may be helpful and easier than using an actual current sensor if the necessary power
supplies or space are not available. A transverse probe can be inserted into the core gap and then a known
current can be passed through a wire going through the core to calibrate the setup for a known
gauss/ampere coefficient. Adding more turns of the current carrying conductor through the core can make a
very sensitive current sensor. This could be useful in situations where trying to measure parasitic current
drain on an otherwise high current circuit. Normally a low current ammeter or shunt used in a situation like
this would be destroyed when the high current was energized. In this case although the core could become
magnetized if the high current were energized and may require degaussing to be accurate at low currents
again, no permanent damage would occur. If the core were not merely gapped, but a 2-piece design similar
to a clamp-on type of current probe, then the core could be removed when the high current was energized
thereby preventing core saturation and magnetization. Reference the figures in our current sensor catalog
for open loop sensors for more details.
