P6.1
Magnetic position sensor with low coercivity material
Jerance N., Frachon D.
Moving Magnet Technologies
1 rue Christiaan Huygens, 25000 Besançon, France
1. Introduction
Magnetic position sensors are widely used, especially in automotive applications. This is due to their low
cost, absence of friction and insensitivity to dirt. Modern magnetic position sensors use measurement of
magnetic field direction [1]. These sensors have a diametrically magnetized magnet which is attached to
the extremity of a rotating shaft, and an ASIC measures the magnetic field angle in the plane below the
shaft.
The rotating part of the sensor is an assembly of two parts: the rotating shaft itself and the magnet. These
materials, in general case, have different thermal expansion coefficients. In order to improve the
mechanical robustness of the sensor and to simplify the production process, we propose to use directly
the shaft (made of steel) already in the application (to guide the rotating part) as the source of magnetic
field or a low coercivity magnetized steel press fitted at the end of the shaft. A particular geometry of the
magnetized part is proposed in order to obtain sufficient induction in the sensor.
In this paper, the properties and the choice of the low coercivity materials are briefly exposed, the
practical realization of prototypes and measurement results are given. Measurements of induction after
temperature cycles and presence of demagnetizing field are presented.
2. Low coercivity materials
As we already said the goal is to use materials having good mechanical properties and significant
magnetization. In certain cases it is difficult to use modern permanent magnets in the automotive
applications because of their mechanical properties such as: fragility, critical use at high temperatures,
difficulty to assemble with mechanical parts. Most of mechanically good materials are not ferromagnetic
at all, or have a very low magnetization. Magnetic steels are a good trade-off: very good mechanical
properties and coercivity (Hc) up to 60-70 Oe [2]. Our choice are carbon steels, which are low cost
materials with sufficient Hc (around 55 Oe). Z30C13 (0.3% C and 13% Cr) is the material which is used
for the measurements and prototypes described below. It is important to mention that the “high” coercive
field is achieved after quenching, as it is shown in Figure 1.
Figure 1: Measured B-H curve for Z30C13 steel before and after quenching
3. Practical realization of the sensor
A new generation of Hall sensors, based on measuring the magnetic field direction, is shown in Figure 2.
These sensors have some advantages over classical Hall sensors, measuring the field strength: no need
to compensate for magnet over temperature, better tolerances, simple structure.
Figure 2: Hall sensor measuring the direction of the magnetic field
In the proposed sensor, the diametrically magnetized magnet is replaced by a ferromagnetic part, having
good mechanical properties , but a coercivity significantly lower than modern permanent magnets.
The main problem when using low coercivity materials is to have enough magnetic field in the system.
The system geometry should be redesigned in order to achieve sufficient induction amplitude for the
probe (few hundreds of gauss). It is possible with a “U” shaped part, as it is shown in Figure 3.
Figure 3: Shape of the ferromagnetic part and its magnetization pattern

Two arms are magnetized in opposite directions, as it is shown by the magnetization vector M . This
shape increases the path length through the magnetized material. We can write Ampere’s law on the path
shown in Figure 3:
H a l a H f l f 0
where:
H a - magnetic field in the air
l a - length of the path through the air
H f - magnetic field in the ferromagnetic material
l f - length of the path through the ferromagnetic material
The value of Ha is intrisic to the selected material, so one can easily notice that the magnetic field in the
air is higher if the path through the magnetized material is longer.
4. Magnetizing method
The ferromagnetic part can be easily magnetized by winding a conductor around one arm of the
ferromagnetic part and by closing the magnetic circuit with another ferromagnetic part, as it is shown in
Figure 4.
Figure 4: Method of magnetizing
The current and number of conductors are calculated by means of the Ampere’s law:
H 1l1 H f l f NI
where:
H1 - magnetic field in the flux path closing part
l1 - length of the path through the flux path closing part
H f - magnetic field in the ferromagnetic material
l f - length of the path through the ferromagnetic material
N - number of conductors
I - magnetizing current
With
H f being equal to the saturation field of the ferromagnetic material, a low coercivity material is
easily magnetized with 500-1000 A current (N is generally low, 1 or 2), for common sensor dimensions.
5. Dimensions of the ferromagnetic part
The value of magnetic field as a function of dimensions and distance to the probe is studied.
We will give a 3D finite element study for three parameters (Figure 5) of a U-shaped ferromagnetic part:
slot thickness, length and diameter of the magnetized part. The results are given for a point located on
the rotation axis at 1 mm below the magnetized part.
Figure 5: Simulations of the induction as a function of the ferromagnetic part dimensions
We notice that the slot thickness is the most important parameter – it should be as low as possible,
however, it is limited by magnetizing conductor thickness. The arms should be as long as possible, and
the diameter as large as possible, but these two parameters are limited by the available space for the
sensor.
6. A measurement example
An example of measurement result for 360° stroke is given in Figure 6 and the dimensions of the
ferromagnetic part are here below:
 E = 2 mm,
 D = 16 mm,
 H = 9 mm.
A measurement of induction is done with 2SA-10 Hall probe [4] (probe able to measured 2 components of
the induction in a single point) and the angle measurement is done with MLX90316 Hall probe [5].
250
200
5
0.5
4.5
0.4
4
0.3
3.5
0.2
3
0.1
induction (G)
100
50
0
-400
-300
-200
-100
voltage (V)
150
2.5
0
2
-0.1
1.5
-0.2
1
-0.3
-150
0.5
-0.4
-200
0
-50
0
-100
-250
100
200
300
400
-0.5
0
60
120
180
240
300
360
angle (deg)
angle (deg)
Figure 6: 360° stroke with ferromagnetic part - measurement result
A good linearity (0.4%) is achieved and the induction at the measurement point is 210 G which is enough
for the considered Hall probe. The axial distance between the ferromagnetic part and the point of
measurement is 1 mm approximately and it is kept identical for all the prototypes mentioned below. For
most automotive applications a shorter stroke is required. Such examples of prototypes are presented in
the next paragraphs.
7. Prototypes for automotive applications
In all prototypes described below the Hall probe MLX90316 [5] is used.
In Figure 7 a prototype of a sensor for an EGR actuator is shown. The sensor part dimensions are E=1.5
mm, H=9.1 mm and D=12 mm. The non-linearity over 80° stroke is given in Figure 7.
0.15
n o n -lin e a rity (% )
0.1
0.05
0
20 0
22 0
240
260
28 0
-0.05
-0.1
-0.15
a ngle (de g)
Figure 7: Sensor prototype for an EGR actuator and non-linearity on 80° stroke
The induction on the measurement point is about 170 G.
A very good non-linearity (close to 0.1 %) is achieved for 80° stroke.
Suspension sensor
A sensor that fits in suspension sensor housing has been made. The dimensions of the ferromagnetic
part are: D = 12 mm, H = 6mm, E = 1 mm. The sensor and the measurement results for 2 strokes (66°
and 86°) are shown in Figure 8.
Figure 8: Suspension sensor: prototype and measurement result
Induction on the measurement point is about 180 G and the non-linearity is below 0.4% for both 66° and
86 ° degree strokes (two possible strokes for this application).
Pedal sensor
For a short stroke application, a pedal sensor has been made. The dimensions of the ferromagnetic part
are E = 1 mm, D = 14 mm, H=10 mm. It is shown in Figure 9, with the measurement result.
Figure 9: Pedal sensor with measurement result
We find about 0.2% non-linearity for 20° stroke with about 250 G on the measurement point. For all
prototypes very good linearity is achieved with a high enough magnetic field for the functioning of the
probe.
8. Stability issues
Stability of magnetization for these sensors should be carefully examined. External field, high
temperatures and ageing can be critical for magnetized steel [2]. In this type of sensor, the induction at
the measurement point should be kept above the critical value for the probe used. A set of tests is done
for stability in temperature, with temperature cycling from -40 to 135 °C (see Figure 10). The maximum
induction in the sensor as a function of the number of temperature cycles is shown in Figure 11.
t em p e ra tu re c y c le s
B afte r te mp e ratu re cycle s
200
13 5
180
11 5
160
140
75
induction (G)
temperature (°C)
95
55
35
15
-5
120
100
B
80
60
0
20
40
60
80
10 0
120
1 40
1 60
18 0
40
200
-2 5
20
-4 5
0
0
tim e (m in )
20
40
60
80
100
120
temperature cycles
Figure 10: Temperature cycles
Figure 11: Measured induction as a function of the
number of temperature cycles
Stability in presence of an external field was also studied. In Figure 12 a measurement result for 40 G
external field applied 10 times is given.
160
140
induction (G)
120
100
80
60
40
20
0
0
1
2
3
4
5
6
7
8
9
10
number of demagnetizings
Figure 12: Change of induction in presence of demagnetizing field
The measured induction decreases after the first external field exposure, after it remains quite stable
(from 134 to 103 G).
9. Sensor in presence of strong external field
After first measurement without external field, a permanent magnet (a cube of approximately 10x10x10
mm and approximately 1T of remanent induction) is brought close to sensor. A measurement is done in
presence of external magnet, then another measurement is done when external magnet is taken away. A
strong magnetic field will disturb any unshielded magnetic sensor, therefore we present only the case
when the magnet is taken away, knowing that the magnet was at 15 mm minimal distance in all
directions. The result is shown in Figure 13.
6
1
5
0.5
Vout
voltage (V)
4
3
0
NL sensor without
magnet
NL magnet removed
2
-0.5
1
0
-60
-40
-20
-1
0
20
40
60
angle (deg)
Figure 13: Influence of strong external field
The ferromagnetic part has kept a sufficient magnetization, but we remark that there is an effect on the
measurement even when the permanent magnet is taken away (the non-linearity is now 0.8% instead of
0.1% with respect to the initial stroke). This is the critical case for this type of sensor and should be taken
into account for the final sensor design.
10. Conclusion
A simple and robust sensor for up to 360° strokes is presented. The low coercivity materials such as
magnetic steels represent a trade-off between magnetic and mechanical properties. With an adaptation of
sensor geometry, the magnetic field in the range of 150 G – 250 G can be obtained, which is sufficient for
the probes which are normally used in magnetic sensors (examples: MLX90316, HMC1512 [6]). The field
amplitude that can be obtained is limited by sensor size. It was shown, however, that this type of sensor
can fit in the sensor housings for the automotive applications, still producing enough magnetic field for the
probe.
Stability of magnetization in presence of external field is examined and tests over wide temperature range
are done. Very good stability in temperature is shown. Moderate external demagnetizing field (40 G, for
example) can decrease the sensor field once, but the magnetization then remains stable if the same
demagnetizing field is applied. Strong external fields (permanent magnets) can demagnetize steel part or
remagnetize it in a wrong direction. Long term stability of the magnetization in such sensors should be
carefully examined.
11. Literature
1. V. Hiligsmann - “360 Degree Rotary Position Sensing with Novel Hall Effect Sensors”, Sensors, March
2006
2. R.M. Bozorth – “Ferromagnetism”, IEEE Press, 1951
3. P.Gandel, D.Frachon, N.Jerance – “Capteur magnétique de position sans contact”, French patent
FR2882432
4. 2SA-10 datasheet, Sentron, www.sentron.ch
5. MLX90316 datasheet, Melexis, www.melexis.com
6. HMC1512 datasheet, Honeywell, www.honeywell.com