DE20210548U1 - Circuit for determining the position of a seat in a motor vehicle has a bridge circuit for varying a bridge voltage with the position of an adjusting device along an adjustment path. - Google Patents

Abstract

Outputs for a bridge circuit connect to a differential amplifier that has an analog amplifying stage and a digital differentiator. A magnetic rod (10) with north (N) and south (S) poles is magnetized parallel to its cross section and crosswise to its axis. A magnetizing section is used for determining positions.

Description

&igr; i.i.. &igr;

Brose Vehicle Parts GmbH & Co.
Limited partnership, Coburg
Ketschendorfer Straße 38 - 50

D-96450 Coburg

BRO1110

Circuit for determining the position of an adjustment device in a Motor vehicle Description

The invention relates to a circuit for determining the position of an adjusting device in a motor vehicle.

In the automotive sector, measurement technology is becoming increasingly important. It is an essential prerequisite for mastering increasingly complex processes. Many customers are placing increasing importance on comfort in their cars. In order to achieve this level of comfort, electrical adjustment options with a memory function on the seat are required, among other things. Operation can be made even easier if it becomes possible to save selected settings and to start again when required, or to make getting in easier by automatically moving the seat (Easy Entry).

BRO1110 : : : :··:·. - : ·'·.:.. Page2

For this position determination, it is known to use a Hall sensor that is contained in the electric motor. This measuring method has a significant disadvantage, since only increments are counted. When the system is first put into operation, a reference run must be carried out in which the two mechanical stops are taught in and their counter readings are saved. This means that the mechanical stops must be designed for these loads.

If the seat is now moved, the counter reading changes according to the direction of movement. The customer moves his seat to a position that is comfortable for him and saves the counter reading for this position. If everything goes smoothly, the actual counter readings are changed when the seat is moved manually, i.e. by using the seat position buttons or by using the memory buttons, and are saved when the seat is stopped. It is also possible to constantly save the actual counter readings; however, the limited number of read and write cycles of the memory modules must be taken into account.

If an error occurs when moving the seat, e.g. an interruption in the power supply or a brief failure of the single-chip microcomputer in the seat control unit, the current actual counter reading can no longer be saved. This means that the seat is mechanically in a different position than the saved counter reading indicates. Due to this deviation, the saved memory positions can no longer be reached precisely and it can happen that the seat is moved back to the mechanical stop. This stop can be permanently damaged as a result.

Since traditionally only a Hall sensor is used for cost reasons, it is not possible to detect the direction of rotation of the motor. The actual counter reading in the seat control unit must therefore be calculated using complex algorithms.

In addition to comfort, safety plays a major role. It is conceivable to adjust the height of the headrest depending on the seating position or to control the deployment of the airbag(s). Another safety aspect is the anti-pinch protection, which has not yet been possible with conventional systems.

BRO1110 : : : :··:··:.:: ·:..:.:.. page3

In seats that can only be adjusted mechanically by hand, there are currently no electronic components for determining the position. Even if a mechanical drive could not be accommodated for reasons of space, but only an electric drive without a Hall sensor and seat control unit is used for manual adjustment for cost reasons, these safety-relevant settings cannot be carried out automatically because the position cannot be determined.

An object of the present invention is to provide a simple circuit which also enables absolute position determination.

This object is achieved by the invention specified in the independent claim. Advantageous embodiments can be found in the subclaims.

According to the invention, a circuit is created for determining the position of an adjusting device in a motor vehicle, comprising: a bridge circuit, wherein the bridge voltage varies depending on the position of the adjusting device along an adjustment path, wherein the outputs of the bridge circuit are connected to a differential amplifier which has an analogue amplification stage and a digital differentiator.

An advantage of this circuit is that no analog reference voltage is required for the measured value acquisition.

In particular, the circuit can perform an absolute length measurement using a magnetoresistive sensor. This means that the seat position is determined absolutely, so that reference runs are not necessary. The mechanical stops on the seat could therefore theoretically be eliminated entirely. Since this measuring device is attached directly to the seat, positioning errors due to changing weight loads on the seat can also be avoided.

With the help of the absolute measuring method using magnetoresistive sensors, it is possible to determine a position without prior teaching of reference points. To do this, it is necessary to align the measuring element (magnetic rod) with the object to be measured. The sensor as a measuring value pickup and the

BRO1110 :..&idigr; ^: .. ^:# T ' ^: .. ^: i " ^: ..·.:.. Page 4

The connected electronics and the software that performs the necessary calculations must be tailored to the measuring task. Of course, all components involved in the measurement must be manufactured and used within the appropriate tolerance in order to display the position with the desired resolution and repeatability.

Preferably, the digital differentiator is integrated in a microcontroller which has one or more analog-digital converters for converting the output signals of the bridge circuit. The microcontroller can be designed to calculate the position of the adjustment device from the superimposed voltages, with the bridge circuit acting as a voltage divider. This solves the problem that bridge elements are not manufactured symmetrically in production and this causes offset problems, ie the output voltages are not symmetrical to half the operating voltage U _b .

The microcontroller can be designed to determine and compensate the offset voltages of the operational amplifiers, the bridge circuit and/or the resistors for amplification. To compensate for the offset voltages, the microcontroller can cyclically determine a reference value. This allows temperature influences that affect the offset voltages to be compensated. This is achieved by calculating an average value of one or more bridge branches. An example formula for this is

U _ref =(U _max -U _min )/2 + U _min
25

The prerequisite for this is that U _max and U _min are (repeatedly) reached.

Alternatively, a separate reference voltage (U _ref ) can be determined for each bridge branch.
30

Furthermore, the microcontroller can be designed to perform a control function for monitoring the bridge circuit and the operational amplifiers, and to calculate the position of the adjustment device based on the calculated offset voltages and a signal from the bridge circuit. This enables a higher

PA105/01

·

BRO1110 :..: :..···..: : ·..= .:.. Page 5

Availability of positioning is guaranteed even in the event of subsystems failure.

Character description

The invention will now be explained using an embodiment with reference to the drawings, in which:

Figure 1 shows a cross section through a magnetic bar;
Figure 2 is a perspective view of the magnetic rod of Figure 1;

Figure 3 shows a relationship between mechanical and calculated position of a seat adjustment with the magnetic rod from Figure 1;

Figure 4 shows a relationship between magnetic field strength and resistance of a magnetoresistive sensor;

Figure 5 shows a magnetoresistive sensor with Barber pole structure;

Figure 6 shows a rotation of the magnetization by an external magnetic field of the magnetoresistive sensor of Figure 5;

Figure 7 the angles of a current and the magnetization in the magnetoresistive

Sensor from Figure 5;
25

Figure 8 shows schematically a magnetoresistive sensor without Barber poles;

Figure 9 shows schematically a magnetoresistive sensor with Barber poles;
Figure 10 shows the basic structure of an angle sensor;

Figure 11 shows the theoretical signal curve of the bridge voltages of a magnetoresistive sensor;

PA105/01

BRO1110 :.. ^{: :} .. ^:# T ' ^: .. ^: i ^#: .. ^: .:.. Page6

Figure 12 shows the real signal curve of the bridge voltages of a magnetoresistive sensor;

Figure 13 shows measured bridge voltages of a magnetoresistive sensor plotted against each other;

Figure 14 shows the circuit diagram of an amplifier circuit for the magnetoresistive sensor;

Figure 14a shows the circuit diagram of Figure 14 with connected microcontroller and adjustment motor;

Figure 15 a direct evaluation of the bridge voltages;
Figure 16 a measuring bridge and its voltages;

Figure 17 shows the basic structure of a differential amplifier with an operational amplifier (OPA);

Figure 18 the direction of the currents in the bridge branch;
20

Figure 19 shows a channel of the measuring amplifier (Pspice simulation);

Figure 20 shows the voltages of the amplifier at a gain V = 25.5;
Figure 21 the voltages of the amplifier at a gain V = 56.4;
Figure 22 the voltages of the amplifier at a gain V = 71;

Figure 23 two bridge signals with offset;
30

Figure 24 ideal signals plotted against each other, with offset;

Figure 25 Angle error caused by an offset of the signals in an ideal

signal curve; and
35

&Oacgr; .\" ^: ":'O :..: j. PA105/01

BRO1110 ·..! \. ^: 'T " ^: ..· : ·..*.:.. Page 7

Figure 26 shows the allocation of the quadrants.

The measuring principle

Figure 1 shows the cross-section of a rod 10 made of plastic-bonded ferrite which serves as a measuring embodiment. This rod is magnetized parallel to the cross-section, i.e. transverse to its axis.

Figure 2 shows a perspective view of the magnetic rod 10. The vector 21 of magnetization rotates over a desired length along the rod axis. The arrows 20 indicate the direction of magnetization.

Also shown is a sensor 22 for detecting the magnetic field. The sensor 22 moves in the axial direction near the magnetic field. Depending on the axial position, a certain vector of the magnetic field is established towards the sensor 22.

The magnetic field vector which is established depending on the position of the sensor 22 relative to the magnetic rod has a different effect on the sensor 22, which results in a resistance change in the bridge resistors of the sensor 22, as described below.

The relationship between the length of the magnetization and the length of the path to be measured is important for calculating the position. The distance between the start of the magnetization and the reference position must also be known. These relationships must either be determined once in the assembled state, or the length and position of the measuring scale must be adjusted to its future use by providing the supplier with the appropriate length information. If the lengths and installation positions remain within a tolerance range and the offsets of the bridge voltages of the magnetoresistive sensor 22 are known (see below), no further measurements are required for adjustment.

The sensor 22 detects the angle of the magnetic field and outputs two voltages. These are amplified and converted to digital values. The position is then calculated from the two values. This position is always absolute to the measuring embodiment.

:.. ^: :... .:.\"&ugr;.:!." I IJ "V·:·:"! : j &Iacgr;&Rgr;&Agr;105/01

BRO1110 :..· ·..· : ·..* : ·..·.:.. Page 8

In this measuring method, no increments are counted. The position is not calculated from the addition or subtraction of increments. It is calculated from two voltages generated by the sensor, which are absolutely related to the position of the sensor in relation to the measuring body. In this case, the measuring body carries the position as an absolute value due to the rotation of the magnetic field. Since the change in resistance in the sensor is accompanied by twice the angle of the magnetic field, the magnetic rod only needs to have a magnetic field rotation of 180° over its length.

Figure 3 shows the relationship between mechanical and calculated position. As can be seen from the illustration, the measuring embodiment can be longer than the adjustment path. However, for optimum position resolution, both should have the same length. If the measuring embodiment were designed for a shorter adjustment path, the position could only be determined absolutely in this shorter range. The position can only be determined absolutely in a signal range of the magnetoresistive sensor from 0° to 360°, i.e. from 0° to 180° magnetic field rotation in the magnetic rod.

The magnetic rod is magnetized over its entire mechanical length during production. Only the required magnetization section is used to determine the position. If this section includes a rotation angle of the magnetic field of less than 180°, clear positioning is possible in any case. However, the value range of the calculated arctangent does not reach 360°. If the range is fully utilized, in this case 0° to 180°, the position can also be determined clearly. If the rotation angle of the magnetization is greater than 180° within the adjustment range, there are at least two mechanical positions over the entire range for one and the same position calculated from the sensor voltages. In this case, sections of the 4 quadrants that form the sine and cosine voltages of the sensor are present multiple times.

To determine the exact position, only the angle range of max. 180° in the magnetic rod should be used for the intended adjustment path. This scaling can be used to assign a position to each angle.

PA105/01

BRO1110 :..· *..· I \.· : ·..*.:.. Page 9

The following calculation is based on returning the read-in voltages, which can be in any quadrant, to the first quadrant. This is possible because the tangent function is a periodic function. If periodic errors occur, they can be corrected in the first quadrant and the position can then be converted to the actual quadrant. These relationships are explained below.

Manufacturing and functionality of the sensor

Magnetoresistive sensors are manufactured using the latest microtechnical manufacturing processes for microchips. They are based on the anisotropic magnetoresistive effect. This effect occurs in certain classes of ferromagnetic materials, such as permalloy. Permalloy is deposited as a thin layer in a magnetic field and structured in strip form. Conductor strips are created with a preset magnetic anisotropy oriented in the longitudinal direction of the conductor during the process.

In the field-free state or when the magnetization and current directions are parallel, this conductor shows a maximum resistance, which decreases when a minimal but sufficiently large external magnetic field is applied. The lowest resistance is reached when the magnetization is perpendicular to the current direction. The maximum change in resistance is a few percent of the total resistance. The change in resistance is determined by the magnetic field strength H and not by the magnetic flux density B. The sensor works in saturation mode above the saturation field strength.

In this structure, the polarity of the field strength plays no role. The change in resistance
is proportional to H ² . However, this does not allow the polarity of the field strength to be determined. By applying Barber poles, which are arranged at an angle of 45° to the longitudinal axis of the sensor, the resistance characteristic of the magnetoresistive sensors is linearized around the operating point. These Barber poles are made of a highly conductive material, e.g. aluminum.

Figure 4 shows the relationship between the magnetic field strength and the resistance of the sensor 22.
35

BRO1110

Figure 5 shows the magnetoresistive sensor 22 with a barber pole structure. The permalloy (approx. 80% nickel, approx. 20% iron) has a very high resistance. Due to the highly conductive metal strips, barber poles, the current I in the field-free space chooses an angle of 45° to the longitudinal direction of the sensor. The current seeks the shortest path through the sensor. If an external magnetic field is now applied to the sensor, the current flow in the sensor and thus the resistance changes depending on the direction of the magnetic field strength. The change in resistance can be calculated as follows.

R(j3) = R _const +AR* cos{2*ß)

Where: R(ß): resistance of the sensor strip

Rcons*: constant resistance component

AR: a few percent of the total resistance

ß: Angle between current direction and magnetization M

Figure 6 shows the rotation of the magnetization M by an external magnetic field. Figure 7 shows the angles of the current direction and the magnetization in the sensor. Figure 8 shows a magnetoresistive sensor without Barber poles; Figure 9 shows a magnetoresistive sensor with Barber poles.

The external magnetic field H rotates the magnetization M and changes the resistance of the magnetoresistive sensor. Since the change in resistance is very small, these sensor resistors are used in bridge circuits. This reduces the influence of temperature and increases sensitivity. The measuring bridges form the basis of the magnetoresistive sensor. The two bridge signals are offset by 90°. Therefore, each magnetoresistive sensor has two of these bridge circuits. The bridges are implemented offset by 45°.

The ideal bridge voltages can be calculated as follows:
30

PA105/01

BRO1110

Where: U _S)N (ß): Bridge voltage of the sine bridge

Ucos(ß): bridge voltage of the cosine bridge

U _A : maximum bridge voltage

Figure 10 shows the basic structure of an angle sensor.

Bridge voltages

The size of the bridge voltages depends on the magnetic field strength H in the magnetoresistive material. Figure 11 shows the theoretical signal curve of the bridge voltages of a magnetoresistive sensor. Figure 12, on the other hand, shows an actual measured curve of the bridge voltages. The offset of both bridge voltages is clearly visible. The signal amplitude and shape also show deviations compared to the theoretical curve. Figure 13 shows the measured bridge voltages plotted against each other.

Bridge amplifier

A direct connection of the magnetoresistive sensor to the single-chip microcomputer used with integrated AD converter is not advisable. The AD converter quantizes its input voltage range from 0 to 5V with a resolution of 10 bits in 1024 increments of 4.88mV each.

Figure 14 shows the circuit diagram of a circuit for amplifying the bridge voltages. Figure 14a shows the circuit with connected microcontroller and adjustment motor.

The bridge voltages of the magnetoresistive sensor are offset and symmetrical to half the operating voltage. Their level is a maximum of 13mV/V. This bridge voltage depends on the operating voltage and the external magnetic field strength H. With an operating voltage of 5V, the sensor delivers a maximum bridge voltage of 65mV. The downstream AD converter only resolves this voltage with 4 bits, and these 4 bits are not used. Due to this low resolution, both bridge voltages are divided into 14 increments.

As mentioned above, the calculation of the position in the 4 quadrants can be reduced to one quadrant. In this quadrant, the two

PA105/01

BRO1110

Voltages that are roughly scaled are used to calculate the position of the sensor. It is conceivable that the calculated position can also only be calculated in this rough grid. Figure 15 shows these relationships. In it, 150 designates the ideal bridge voltage of the sine bridge, 151 the ideal bridge voltage of the cosine bridge, 152 the quantized bridge voltage of the sine bridge, 153 the quantized bridge voltage of the cosine bridge, and 154 the calculated position in the first quadrant.

The rough grid of the bridge voltage can be clearly seen. This representation is based on ideal sine and cosine voltages. The angle was calculated using the arctangent function. Using appropriate scaling, a position can be assigned to this angular range of 90° and, converted to all 4 quadrants, to an angular range of 360°.

The calculated angle can only be read in a rough grid of different grid widths. Assuming an ideal step function, the quadrant can be divided into 20 steps. This corresponds to a resolution of about 22% of the adjustment length. A higher resolution is achieved by using the entire converter width. In this illustration, it is clear that no clear assignment of the true and calculated angle is possible. Operating this arrangement is therefore not sufficient for a required resolution of <=1%. The use of an additional amplifier is unavoidable.

In order to be able to use the AD converter in its entire input range, it is necessary to amplify the bridge voltages and adapt them to the AD converter input. The requirements for the circuit were defined as follows:

1. Use in the automotive sector

2. Amplification of bridge voltages

3. Supply of an AD converter with a maximum input voltage of 5V

4. Unipolar power supply

5. Voltage resistant according to the requirements in the vehicle

6. Temperature range -40 ⁰ C to 8O ⁰ C

7. Cost-effective option

8. No additional AD converter required, as built into microcontroller

PA105/01

BRO1110

The magnetoresistive sensor works with a supply voltage of up to 8V. Input voltages greater than 5V are not permitted for the downstream AD converter. In order to avoid additional circuits for level adjustment, the sensor and amplifier circuit are operated with 5V operating voltage. This guarantees protection of the AD converter input. This operating voltage is sufficient for the magnetoresistive sensor.

Since a voltage regulator with the necessary external circuitry that stabilizes and smoothes the operating voltage is required later to supply the single-chip microcomputer, additional components are not required. The bridge voltages generated in the magnetoresistive sensor refer to half the operating voltage. A maximum level of ±65mV must be expected between the sensor outputs. This level should be amplified to ±2.5V. In this case, the input voltage range of the AD converter could be fully utilized.

Since offset voltages are added to the bridge voltages of the sensor and the bridge voltages also depend on the magnetic field strength H, the gain of the bridge amplifiers cannot be set to maximum. It must be guaranteed that the operating range of the amplifier circuit is not exceeded in any situation. The amplifiers are limited by the operating voltage. In order to be able to use as large a range as possible within the operating voltage limits, components must be used that work up to a few millivolts of the operating voltage range. These are operational amplifiers with rail-to-rail behavior at the output.

The maximum possible gain can be calculated as follows:

r. ^u -

u _Br

Where: V: reinforcement

U _AD : Voltage at the AD converter input

U _Br : Bridge voltage of the magnetoresistive sensor

PA105/01

Please note that only the bridge voltage is amplified.

Figure 16 shows a measuring bridge and its voltages.
5

Ideally, the outputs of the measuring bridge have the following voltage levels:

" 2
10

The voltage AU is generated in the magnetoresistive sensor by moving it on the magnetic rod. The resistors in the bridge optimally change their values by the same amount. The bridge voltage to be tapped is calculated as follows:

U _Br =U _el -U _el =±2AU

Since it is not possible to operate the amplifier circuit with a symmetrical operating voltage, the amplified bridge voltage must be shifted by an offset of half the operating voltage. In this way, the output of the amplifier circuit can be controlled around its operating point.

Figure 17 shows the basic structure of a differential amplifier with an OPV according to an example. The pure differential amplifier is used without an additional OPV. Due to this saving, one branch of the measuring bridge is loaded, since this branch is connected to the inverting input of the amplifier circuit. The resistors for adjusting the gain must be selected so that the current through this one measuring branch, starting from the OPV, is as low as possible. However, this is not easily possible, since only resistors up to a maximum value, about 18OkΩ, are used.

The voltage U _e1 is provided by one branch of the measuring bridge. If the operating voltage is applied to U _e2 , R ₇ and R ₈ form the other branch of the measuring bridge.

PA105/01

BRO1110

The following relationships also apply:

^K6

Depending on U ₀ and U _e2 the direction of the current I _R6 varies.

In this bridge branch, the following relationships apply in the unloaded case:

¹ Br

U _el =I _Br *R ₄

Normally, the measuring bridge is not loaded by the connected electronics. In this case, the current I _R6 acts on the bridge. In order to keep this reaction as low as possible, a particularly high resistance value must be used for R ₅ and R ₆ .
Figure 18 shows the direction of the currents in the bridge branch.

Figure 19 shows one channel of the measuring amplifier (Pspice simulation).

For better common mode rejection, the future use of resistors R ₇ and R ₈ makes sense.

Figures 20, 21 and 22 show the voltages at a gain of V=25.5, V=56.4 and V=71 respectively.

Due to the excessively high gain in Figure 22, the amplifier is overloaded. In these areas, a deviation from the calculated position is to be expected.

In general, however, the bridge voltage can be amplified to a high degree. This results from the influence of the operational amplifier on the bridge voltage.
30

The influence of the individual offset voltages of the measuring bridges is present. With the bridge voltages calculated in this way, the position can be determined with deviation in linearity

PA105/01

BRO1110 :..· ·..· : ·..· :

If the linearity error resulting from this method is too large, the two offset voltages must be determined and compensated.

The bridge voltages are composed of the following values: 5

Tj -Tj ₊ y*rj

^U AD-sin - ^U Off-sin ^ ^{Y U} sin-Br
U AD-cos ⁼ U Off-cos ⁺ " ^cos-flr

The two bridge voltages are now calculated as follows: 10

SINUS_0 = U _AD _ _sin

COSINE _O = U ^ ₀₀₅

Offset correction

The aim of the offset correction is to center the offset bridge voltages to the abscissa. This achieves a linearization of the position calculation.

The following procedures can be used for offset correction:

1.) For each measuring device individually:

One-time determination of the maxima of the bridge voltages after assembly

the measuring device
- Saving the 4 maxima

Calculating the two offsets
- Correct bridge voltages

2.) The same for each measuring device:

For each batch Determination of the 4 maxima in a measurement setup Calculation of the offsets

- Entering the offsets when programming the single-chip microcomputer for the measuring device

BRO1110

In the present case, method 2 was used for offset correction. The offset-corrected values (SINUS and COSINE) were available for further calculations.

Figure 23 shows two ideal bridge signals, each shifted by an offset. In order to better recognize and determine the offset, it makes sense to display the signals relative to each other. After determining the 4 maxima, the offset shown can be calculated and the signals corrected. The curve shown is then placed symmetrically to the zero point. The stability of the corrected values in the temperature and operating voltage range may still need to be determined. If the offset changes under changing ambient conditions, measurement errors must be expected.

Figure 24 shows ideal signals plotted against each other; the offset is also entered. 15

Figure 25 shows the angle error caused by an offset of the signals in an ideal signal curve.

For an offset of both bridge voltages of 10% of the maximum signal amplitude, Figure 24 shows the angle error of the calculated angle. Since this error has no influence on the repeatability, but only on the linearity of the calculated angle, it must be decided whether it is necessary to determine the offset. It has been found that it may be sufficient to determine the offset with only a low resolution.

Calculating the position

The starting point for further calculations are the two bridge voltages.

Figure 26 shows the assignment of the quadrants.
The position is calculated using the following steps:

1. Determining the quadrant and storing the quadrant number

2. Calculation of the bridge values

3. SINUS = 0? then intermediate position = 0°

4. COSINE = 0? then intermediate position = 90°

BRO1110 i.. ^: \. ^: "&iacgr; - ·--·

5. For all other values, calculate the quotient of SINUS to COSINE

6. Find the appropriate value from the table (tangent from 0° to 90°)

7. output appropriate angle to the calculated value

8. Return the calculated intermediate position to absolute position 5

The following relationships apply to the 4 quadrants in the angle range 0° to 360°:

I.Quadrant: Bridge value SINUS: positive

Bridge value COSINE: positive 10

Absolute position = intermediate position

2nd quadrant: Bridge value SINUS: positive

Bridge value COSINE: negative 15

Absolute position = 180° - intermediate position

3. Quadrant: Bridge value SINUS: negative

Bridge value COSINE: negative 20

Absolute position = 180° + intermediate position

4th quadrant: Bridge value SINUS: negative

Bridge value COSINE: positive 25

Absolute position = 360° - intermediate position

About the equation

360°

the length position can be calculated from the absolute position and a length can be assigned to the angle, where

PA105/01

BRO1110 :.. ^: \. ^: *&Tgr; * ^: .. ^: :

β: Angle between current direction and magnetization M

I: maximum adjustment range

S: calculated position

Tangent table

In order to limit the computing time, it is necessary to work only with integers.

It is therefore necessary to work with the help of a table that contains the necessary number of tangent values in the angle range 0° to 90° in an appropriate form. As already mentioned, the calculation can be carried out in the first quadrant. This reduces the length of the table and thus the memory required.

The required resolution is a maximum of 1% of the total length. For a total angle of 360°, this corresponds to a minimum resolution of 3.6°. The aim should not only be to achieve this resolution, but also to evaluate the measuring device. To do this, the angle range from 0° to 90° is divided into 0.25° steps. The table therefore contains 361 values. This method can be used to limit the number of rows. The possible number of calculated tangent values by forming a quotient from the read-in SINE and COSINE values would exceed the memory size.

An angle is not assigned to every possible calculated tangent, but rather, in order to achieve the resolution, a tangent value is assigned only to the necessary angles. As stated in the above equations, the positions in the individual quadrants are calculated. For this purpose, only the value 180° corresponds to a value of 720 and the value 360° to a value of 1440. The correspondingly scaled values of the tangent in the specified angle range are assigned the numbers 0 to 360. This table is intended in particular in connection with reading in the voltages of the 4 bridge branches. Each of these 4 values is converted with 10 bits. This results in 4 numbers in the range 0 to 1023. The bridge values for sine and cosine are calculated using the following formulas:

SINE _- O = U ₄₀ ^ ₁ -U _AD _ _A2 = U _off _ _sin+ V*U _sin _ _Br COSINE _ O = £/^_ _cosl - U ₄₀ ^ ₂ = U _off _ _ms + V * U _C0S &ldquor; _Br

9· · 9 9 9 999 9 9*99 9 9 9 9 9

·· 9999 999 9999 999

9 9999 9999 99 9999 999 99 999 9 99 99 999

PA105/01

BRO1110

These bridge values therefore theoretically cover the entire value range.

The aim is to calculate as many usable integer TANGENT values as possible when dividing the SINE and COSINE values. Poles and zeros are excluded during the calculation. The following calculation rule is chosen for this purpose

TANGENT =

SINUS* 1000
COSINE

According to this equation, the single-chip microcomputer calculates the tangent from the read and partially offset-corrected values, which is then compared with the table values.

The values in the tangent table contain the same set of values as the calculated tangent value from the read data.

The values in the table are calculated using the following formula:

ROUND(1000*TAN(ANGLE))

20 Where:

ANGLE: Angle range 0° to 90°, step size 0.25°

The ROUND command rounds the values to whole numbers.

25

This results in the following tangent table:

Line number Angle in ° TAN(angle) 0 0 0 1 0.25 4 2 0.5 9 3 0.75 13 4 1 17 356 89 57290 357 89.25 76390

BRO1110

358 89.5 114589 359 89.75 229182 360 90 1023001

Using the values calculated in this way, results are obtained in the specified
tolerance range achieved.

Perhaps an increase in the step size, e.g. to 1 degree, may be sufficient for the
required resolution. This significantly reduces the memory requirements.

It should be noted that the present invention is not limited to the embodiment described, but includes modifications within the scope of protection defined by the claims.

·

BRO1110

List of reference symbols

1 Measuring standard with specified magnetic field direction rotation 180° 2 Bridge voltages generated by the sensor, each one full period 3 Bridge voltage of the cosine measuring bridge 4 Bridge voltage of the sine bridge 5 calculated arctangent from the values given at the respective position Bridge voltages 6 MR sensor without barber pole 7 MR sensor with Barber poles 8th Permalloy 9 Barber Pole 10 Magnetic bar 11 MR sensor 12 Ua(sin) 13 Ua(cos) 14 Microcontroller 15 drive 16 Ir ₆ for U ₀ <Ue2 ■* U _e1 *<Uei for Ir ₆ =O 17 Ir ₆ for U ₀ >Ue2 ■*· Uei*>Uei for l _R6 =0 18 0.5+sin(angle) 19 0.5+cos(angle) 20 Magnetization direction 21 Magentization vector 22 magnetoresistive sensor 23 Offset 24 calculated angle 25 ideal angle 26 Angular error 27 Sine 28 Cosine

&bull; t · ·

&bull; r &bgr;··· *

BR01110

Page 22a

150 ideal bridge voltage of the sine bridge

151 ideal bridge voltage of the cosine bridge

152 quantized bridge voltage of the sine bridge

153 quantized bridge voltage of the cosine bridge

154 calculated position in the first quadrant Ub operating voltage

U _e i Output voltage 1

Ue2 Output voltage 2 Ri - R ₄ magnetoresistive resistors

Claims (18)

1. Circuit for determining the position of an adjusting device in a motor vehicle, comprising:
a bridge circuit, wherein the bridge voltage varies depending on the position of the adjustment device along an adjustment path,
characterized in that
the outputs of the bridge circuit are connected to a differential amplifier which has an analog gain stage and a digital differentiator. 2. Circuit according to claim 1, characterized in that the analog amplification stage consists of two coupled operational amplifiers. 3. Circuit according to claim 2, characterized in that the inputs of the operational amplifiers for the feedback voltages are connected to one another via a resistor. 4. Circuit according to one of the preceding claims, characterized in that the digital differentiator is integrated in a microcontroller which has one or more analog-digital converters for converting the output signals of the bridge circuit. 5. Circuit according to claim 4, characterized in that the microcontroller is designed to calculate the position of the adjusting device from the superimposed voltages, wherein the bridge circuit acts as a voltage divider. 6. Circuit according to claim 2 or 3 and claim 4 or 5, characterized by a multiplexer (MUX) for cyclically connecting the analog-digital converter(s) to one of the outputs of the operational amplifiers. 7. Circuit according to claims 2 and 4, characterized in that the microcontroller is designed to determine and compensate the offset voltages of the operational amplifiers, the bridge circuit and/or the resistors for amplification. 8. Circuit according to claim 7, characterized in that the microcontroller is designed to compensate the offset voltages for the cyclic determination of a reference value. 9. Circuit according to claim 7 or 8, characterized in that the microcontroller is designed to carry out a control function for monitoring the bridge circuit and the operational amplifiers, and for calculating the position of the adjusting device based on the calculated offset voltages and a signal from the bridge circuit. 10. Circuit according to one of claims 4 to 9, characterized in that the microcontroller is designed to calculate the position of the adjusting device by processing the digitized signals to form a tangent function. 11. Circuit according to claim 11, characterized in that the processing of the digitized signals comprises the formation of an angle value by assignment to a tangent value calculated on the basis of the tangent function within a predetermined quadrant. 12. Circuit according to claim 12, characterized by a memory for storing a plurality of angle values within the angle range of the predetermined quadrant, wherein the microcontroller is designed to retrieve the angle value associated with the calculated tangent value. 13. Circuit according to claim 11 or 12, characterized in that the predetermined quadrant is the first quadrant. 14. Circuit according to one of claims 11 to 13, characterized in that the processing of the digitized signals comprises the determination of a quadrant, and the microcontroller is designed to calculate the position of the adjusting device on the basis of the quadrant determined in this way and the angle value determined within the predetermined quadrant. 15. Circuit according to claim 14, characterized in that the microcontroller is designed to determine the quadrant based on the signs of the bridge voltages. 16. Circuit according to one of claims 4 to 15, characterized in that the microcontroller is designed to linearize the position calculation for carrying out an offset correction of the digitized signals on the basis of stored offset values. 17. Circuit according to one of the preceding claims, characterized in that the bridge circuit is designed such that its output signals are phase-shifted by approximately 90° to one another. 18. Circuit according to one of the preceding claims, characterized in that the adjusting device is an electric seat adjustment.