Third-Party Research & Development
Technical Discourse
Extracting Static and Dynamic Information From
Hallpot® Angle Sensors:Linear & Sinusoidal Sensors
John
R. Potts, Senior Scientist/Engineering Manager
BRIEF
Elweco Inc. (Painesville, OH) manufactures and markets rotary transducers
based on the Hall Effect. Essentially, two basic types of calibrated
transducers are available: (1) directly coupled Hallpot® Angle
Sensors, for precise system control and/or measurements of angular
displacements generally over restricted ranges (+/- 30 degrees) and
(2) directly coupled Sin-Cos Hallpot® Resolvers, for precise system
control and/or measurements of angular displacements generally over
unrestricted ranges (+/- 360 degrees or more). This paper discusses
methods for determining shaft position, calibrating sensor output
over a restricted angular range, determining shaft speed, determining
shaft acceleration, and an application for extracting mechanical properties
of relatively soft materials, using a Hallpot® Angle Sensor. For
this paper, isothermal behavior is assumed. Thermal effects are discussed
in Temperature Behavior of Hallpot®
Angle Sensors and Sin-COs Hallpot® Resolvers, published
on this website. Techniques for extracting information from Sin-COs
Hallpot® Resolvers are discussed in Extracting
Static and Dynamic Information From Sin-COs Hallpot® Resolvers:
The ATAN2 Function.
Third-Party Research & Development fabricates and
markets diagnostic systems based on RVDT and HallPot® rotary sensors.
IDRA® Ballistometer System, which is used in cosmetic claims testing,
uses a HallPot® Angle sensor and dynamically assesses and is applied
to monitor changes in several mechanical properties of skin and other
soft materials. Third-Party R&D also provides custom hardware
and software engineering development support services for HallPot®
sensors.
TYPES OF HALLPOT® ANGLE SENSORS
Elweco, Inc. provides custom designed Hallpot® Angle Sensors.
Two major types of sensors are available depending on the relationship
desired between sensor output and shaft angle: (1) Linear and (2)
Sinusoidal. They differ in terms of the kind of internal amplifier
circuitry employed, type of output waveform, and range of useful
operation. With appropriate techniques, both kinds of sensors are
capable of providing excellent linearity and reproducibility over
restricted ranges (e.g. +/- 30 degrees). A Linear Hallpot® Angle
Sensor is factory calibrated to user specifications and generally
requires a straightforward calibration procedure to put it into
service. Sinusoidal sensors generally require a more involved calibration
procedure, however, offer a wider range of operation. Figure 1 is
a typical waveform obtained with a Linear Hallpot® Angle Sensor
and shows two linear regions, each spanning nearly 60 degrees. Figure
2 is a typical waveform obtained with a Sinusoidal Hallpot®
Angle Sensor and shows the output for one revolution of the shaft.
Given appropriate technique, both types of Hallpot® Angle Sensors
are capable of resolving an absolute angular position, over a restricted
measurement range, to +/- 0.1 degree or better.
Figure 1- Linear HallPot® Angle Sensor

Figure 2- Sinusoidal HallPot® Angle Sensor
(1) LINEAR HALLPOT® ANGLE SENSOR
CALIBRATION PROCEDURE
At constant temperature, the sensor output, over restricted angular
ranges, is linear with shaft angle:
Vlin=Zero+Sensitivity*(angle) [1]
Where, Vlin is the output response of a linear angle
sensor at a particular shaft position in volts, Zero is a factory
adjusted offset, in volts, and Sensitivity, in volts/degree, is a
slope that is factory adjusted according to user specifications. Sensitivity
is positive and sensor output increases with counterclockwise shaft
rotation in the lower linear angular range, whereas Sensitivity is
negative and sensor output decreases with counterclockwise shaft rotation
in the upper linear angular range (see Figure 1). A protractor or
digital clinometer and DC meter with 4 digit readout set to read 0-10
volts, mounting plate to support the sensor, and pointer arm which
can be attached to the sensor shaft to allow angular measurements,
are recommended for checking and/or calibrating the Zero and Sensitivity
values for a sensor. The calibration procedure simply records the
sensor output at two known shaft positions within the linear range.
Sensitivity is then calculated as the difference in the readings divided
by the known angular difference. The Zero for a sensor is calculated
using the expression Vlin-Sensitivity*(angle).
ALIGNMENT PROCEDURE
In equation [1], the measurable value of Zero offset can differ from
the factory set value. It is user adjustable and can be changed by
holding the shaft at a fixed position while rotating the sensor housing.
Changing the value of Zero for the sensor, in this manner, allows
angular control and/or measurement about an arbitrary reference position.
For example, the positive Sensitivity of the sensor waveform shown
in Figure 1 is 0.1667 volt/degree and the factory Zero is 0.012 volt.
Holding the shaft and rotating the sensor housing until the sensor
output reads 1.667 volts, sets the value of Zero in [1] to that voltage
value and in effect moves the reference angle from 0.00 to 10.0 degrees.
Angles calculated using [1] would indicate shaft positions ranging
from -10 degrees to + 50 degrees with respect to the reference position.
DETERMINING SHAFT SPEED FOR A LINEAR HALLPOT®
ANGLE SENSOR
Accurate real-time determinations of the shaft speed of a rotating
angle sensor requires that the sensor output voltages are sampled
at an adequate rate. Within the sampling time interval, (Vlin-Zero)/Sensitivity
must be evaluated to determine an instantaneous shaft position, i.e.,
absolute shaft angular displacement at the particular sampling time.
A change in the determined angular displacement, for successive samples,
divided by the sampling time interval is a measure of the angular
speed during that interval. If the sampling time interval is sufficiently
small, then the angular speed calculated will be an accurate approximation
of the instantaneous shaft speed.
DETERMINING SHAFT ACCELERATION FOR
A LINEAR HALLPOT® ANGLE SENSOR
Determining shaft acceleration is similar to the problem of determining
shaft speed, since it also requires adequate data sampling and an
accurate estimate of shaft position. If the sampling time interval
is sufficiently small, then calculation of the acceleration, i.e.,
the change in angular speed between two successive time intervals
divided by the time interval, will be an accurate approximation of
the instantaneous shaft acceleration.
(2) SINUSOIDAL HALLPOT® ANGLE
SENSOR
CALIBRATION PROCEDURE
At constant temperature, the sensor output follows biased sine wave
behavior:Vsin=Ebo+Vp*sin(angle) [2]where, Vsin is the output voltage
response of the angle sensor at a particular shaft position , Ebo
is a constant offset voltage with magnitude nominally equal to half
the supply voltage (Vcc), VP is the maximum voltage change in Vsin
due to shaft rotation (generally less than half the supply voltage),
angle is the absolute angular displacement of the shaft in the rotational
plane of the sensor system, and sin is the familiar transcendental
mathematical function which describes precisely how the sensor output
changes with the shaft angle.
Ebo and VP are generally known for calibrated sensors. If the offset
and slope are unknown, they can be easily determined. Averaging the
minimum and maximum sensor output voltages provides a reasonable estimate
for Ebo. An estimate for VP is provided by dividing the difference
between the respective minimum and maximum output voltages by 2. Overall,
the best agreement between the observed sensor outputs and those calculated
using [2] is achieved with an Ebo and VP determined from a least squares
fit of Hallpot® Angle Sensor outputs to the basic defining equation
[2]. The arcsine is a transcendental mathematical function which provides
linear analogs of angular displacements over restricted ranges. Absolute
shaft position is accurately calculated using the arcsine function,
which is evaluated from sensor outputs using [3]:
ASIN ((Vsin-Ebo)/VP) [3]
ALIGNMENT PROCEDURE
The sensor housing can be rotated, with the shaft held at a fixed
position in the rotational plane, to align a reference angle to that
position.
DETERMINING SHAFT SPEED
Accurate real-time determinations of the shaft speed of a rotating
Sinusoidal HallPot® Angle Sensor requires that output voltages
are sampled at an adequate rate. Within the sampling time interval,
the asin function must also be evaluated and used to calculate an
instantaneous shaft position, i.e., absolute shaft angular displacement
at the particular sampling time. A change in the determined angular
displacement, for successive samples, divided by the sampling time
interval is a measure of the angular speed during that interval. If
the sampling time interval is sufficiently small, then the angular
speed calculated will be an accurate approximation of the instantaneous
shaft speed.
DETERMINING SHAFT ACCELERATION
Determining shaft acceleration is similar to the problem of determining
shaft speed, since it also requires adequate data sampling and an
accurate estimate of shaft position. If the sampling time interval
is sufficiently small, then calculation of the acceleration, i.e.,
the change in angular speed between two successive time intervals
divided by the time interval, will be an accurate approximation of
the instantaneous shaft acceleration.
AN APPLICATION - IDRA® Ballistometer
IDRA (Integrated Dynamic Rebound Analyzer) is a turnkey, PC-based
instrument manufactured and marketed by Third-Party Research &
Development, which employs a dynamic technique for assessing the intrinsic
viscoelastic properties of a material. See Figure 3, which is a block
diagram of the IDRA system. For relatively soft materials, such as
skin, the measurement technique involves a lightweight hammer, anchored
at one end to a rotary transducer, which free-falls onto the test
surface under gravitational force, recording and analysis of the resulting
hammer oscillatory displacement-time data, and the determination of
characteristic physical parameters. This instrument accepts either
an RVDT or Hallpot® Angle Sensor, as the rotary transducer. Comparable,
reliable performance is found for these types of sensors. Hallpot®
sensors are recommended for client systems because they offer design
simplicity and significant cost savings.
. 
Figure 3 - IDRA® Ballistometer
System
Figure 4 shows the kind of mechanical diagnostic information extracted
with IDRA® Ballistometer and a Hallpot® Angle Sensor. The
parameters listed are determined for adjacent sections of successive
curve peaks and valleys in a basic response curve. The IDRA HELP menu
shows techniques for extracting additional, detailed information from
this data.
. 
Figure 4 - Mechanical Diagnostic Information Provided
By IDRA® System
Acknowledgment:
Data shown in Figure 2 was provided by John Wereb of Elweco, Inc.
IDRA® is registered to John Potts and Third-Party Research &
Development, Hallpot® Angle Sensor and.Sin-Cos
Hallpot® Resolver are registered to Elweco, Inc.