PROXIMITY SENSOR In the accompanying drawings: FIG. la illustrates a proximity sensor having three control electrodes, and a block diagram of an associated circuit; FIG. lb illustrates various control states of the proximity sensor of Fig. la; FIG. 2 is a contour plot of the electric field potential for a three plate capacitor in XY Space, for an in-phase (IP) mode of operation; FIG. 3 is a plot of the Y-component of the electric field as a function of position along the sense electrode surface of a three plate capacitor for the in-phase (IP) mode of operation illustrated in Fig. 2; FIG. 4 is a contour plot of the electric field potential for a three plate capacitor in XY Space, for an out-of-phase (OOP) mode of operation; FIG. 5 is a plot of the Y-component of the electric field as a function of position along the sense electrode surface of a three plate capacitor for the OOP mode of operation illustrated in Fig. 4; FIG. 6 is a comparison of the capacitance of a proximity sensor as a function of target distance for various modes of operation; FIGs. 7a-f illustrate various embodiments for switching a calibration capacitor; FIG. 8 illustrates a flow chart of the operation a proximity sensor; FIG. 9 illustrates various regions of a DeltaCap measure; FIG. 10a illustrates the effect of a dielectric on a proximity sensor operated in an out- of-phase (OP) control mode; FIG. 10b illustrates the effect of a dielectric on a proximity sensor operated in an in- phase (IP) control mode; FIG. 11 illustrates the effect of a dielectric on the capacitance of a proximity sensor as a function of distance for an out-of-phase (OP) control mode, with and without an in-phase (IP) guard;

FIG. 12 illustrates the effect of an in-phase (IP) guard on the capacitance of a proximity sensor for various dielectric objects; FIG. 13 illustrates various measures for targets moving closer to and farther from a proximity sensor; FIG. 14 illustrates various measures for a dielectric target moving away from a proximity sensor; and FIG. 15 illustrates a front-end amplifier with an offset canceling circuit in accordance with the instant invention.

Referring to Fig. la, a proximity sensor 10 comprises a conductive sense electrode 12 --a first electrode--and at least one conductive control electrode 14--at least one second electrode--for example a plurality of control electrodes 14.1,14.2 and 14.3, separated from the sense electrode 12, so that an electric field 16 created by the sense electrode 12 and at least one of the control electrodes 14.1,14.2 and 14.3 occupies a region of space 18 within which a proximity of an electric-field-influencing media 20 is sensed. The sense 12 and control electrodes 14.1,14.2 and 14.3 are, for example, disposed on a common, surface 22, wherein the control electrodes 14.1,14.2 and 14.3 are disposed outside a periphery 24 of the sense electrode 12, are separated from one another, and at least partially surround the sense electrode 12.

A first oscillatory signal 26 is applied to the sense electrode 12 by a signal generator 28. A second oscillatory signal 3 is applied by a control electrode driver 32 to the control electrodes 14.1,14.2 and 14.3, whereby a phase of the second oscillatory signal 30 relative to the first oscillatory signal 26 is controlled by a controller 34. Referring to Fig. lb, wherein the"+"and"-"signs indicate signals that are respectively in-phase and out-of-phase with respect to the first oscillatory signal 26--the second oscillatory signal 30 comprises a state 36 corresponding to the phase of the second oscillatory signal 30 relative to the first oscillatory signal 26, and the controller 34 controls this state. When in a first state 38 the second oscillatory signal 30 has a first phase relative to the first oscillatory signal 26, when in a second state 40 the second oscillatory signal 30 has a second phase relative to the first oscillatory signal 26. For example, with a sinusoidal first oscillatory signal 26, in the first state 38, the first phase is substantially one hundred eighty (180) degrees so that the second oscillatory signal 30 is substantially out-of-phase

with the first oscillatory signal 26, and in the second state 40, the second phase is substantially zero (0) degrees so that the second oscillatory signal 30 is substantially in- phase with the first oscillatory signal 26.

A first circuit 42 operatively coupled to the sense electrode 12 senses a response signal 44 from the sense electrode 12. The response signal 44 is responsive to the first 26 and second 30 oscillatory signals, and to a proximity of an electric-field-influencing media 20 to the sense electrode 12. The response signal 44 is, for example, a current from the sense electrode 12 responsive to applied voltages of the first 26 and second 30 oscillatory signals, or a voltage responsive to the current. Generally, the response signal 44 either is used to determine the capacitance of the sense electrode 12, or the response signal 44 is a measure of the capacitance of the sense electrode 12.

For example, in the circuit 36 illustrated in Fig. la, the first oscillatory signal 26 comprises a voltage applied to a voltage divider 46 comprising a capacitor Cl and the first electrode 12 wherein one terminal of the capacitor Cl is operatively coupled to the first electrode 12 at a first node 48 and another terminal of the capacitor Cl is operatively coupled to the first oscillatory signal 26, wherein a displacement current flows through capacitor Cl and the first electrode 12. The resulting voltage drop across the first electrode 12 at the first node 48 is sensed as the response signal 44. More particularly, when the second oscillatory signal 30 is in the first state 38, a third signal 50 is sensed at the first node 48 responsive to an out-of-phase signal applied to the control electrodes 14.1,14.2 and 14.3 and when the second oscillatory signal 30 is in the second state 40, a fourth signal 52 is sensed at the first node 48 responsive to an out-of-phase signal applied to the control electrodes 14.1,14.2 and 14.3.

In the exemplary system illustrated in Fig. la the signal generator 28 generates a continuous wave sinusoidal signal having a frequency of 100 KHz. The signal generator is operatively coupled to a front end amplifier 54 comprising the voltage divider 46 and a buffer amplifier Ul. A buffered first oscillatory signal 56--a"buffered"version of the first oscillatory signal 26--is operatively coupled to the control electrode driver 32, which generates the second oscillatory signal 30 that is either in-phase or out-of-phase with respect to the first oscillatory signal 26, under digital control of the controller 34.

The output 58 of the front end amplifier 54--at the same frequency as the signal

generator 28--increases in amplitude as the capacitance of the sense electrode 12, or that of an associated parasitic capacitance, increases. Accordingly, the output 58 of the front end amplifier 54 provides a measure of the capacitance that is operatively coupled to the first node 48. Furthermore, when the second oscillatory signal 30 is in the first state 38 so that the control electrodes 14.1,14.2 and 14.3 are driven out-of-phase (OP) with respect to the sense electrode 12, the associated third signal 50 at the output 58 provides a measure of out-of-phase capacitance Cop of the sense electrode 12. Moreover, when the second oscillatory signal 30 is in the second state 40 so that the control electrodes 14.1,14.2 and 14.3 are driven in-phase (IP) with respect to the sense electrode 12, the associated fourth signal 52 at the output 58 provides a measure of in-phase capacitance Clp of the sense electrode 12. The front end amplifier 54 incorporates a second circuit 60--described more fully hereinbelow--that partially cancels the associated offset capacitance COFFSET SO as to enable a higher gain in the front end amplifier 54, without saturation which could otherwise result without this cancellation.

The front end amplifier 54 also incorporates a calibration capacitor CCAL that can be operatively coupled in parallel with the sense electrode 12 by activating FET transistor Qla and deactivating FET transistor Q2a for purposes of calibrating the first circuit 42, for example to compensate for drift of component values over time or due to temperature variation. The calibration capacitor CCAL can be decoupled from the sense electrode 12 by deactivating FET transistor Qla and activating FET transistor Q2a. As will be described more fully hereinbelow, this arrangement provides a means for coupling and decoupling a capacitor to the sense electrode 12 without introducing associated parasitic capacitances of the associated switch elements to the first node 48. The associated states of the FET transistors Qla and Q2a are mutually exclusive, under digital control, for example by the controller 34.

Similarly, the front end amplifier 54 also incorporates a bias capacitor COTAS that can be operatively coupled in parallel with the sense electrode 12 by activating FET transistor Qlb and deactivating FET transistor Q2b for purposes of augmenting the in- phase capacitance Clp so that the associated third signal 50 is of similar magnitude to the fourth signal 52. This is useful, because as will be described more fully hereinbelow, the out-of-phase capacitance Cop is generally of greater magnitude than the in-phase capacitance Clp, so by making the net capacitance at the first node 48 similar in both

cases, both measurements can be made with the same amplifier gain so as to maximize dynamic range. The bias capacitor CBIAS can be decoupled from the sense electrode 12 by deactivating FET transistor Qlb and activating FET transistor Q2b.

A high pass filter 62 operatively coupled to the output 58 of the front end amplifier 54 filters out base band signals (particularly 60Hz) in the output 58, before detection. This is useful because the detector 64 is essentially a peak detector and cannot distinguish between the envelope signal around the signal at the frequency of the signal generator 28 and additive signals at the base band. In the exemplary system, the high pass filter 62 comprises a 2-pole filter with a 10 KHz cut-off frequency.

The detector 64 is operatively coupled to the output of the high pass filter 62, and, for example, comprises a simple diode detector, which is the principal non-linear factor in the circuit. Accordingly, the drive into the detector is preferably maintained at a constant amplitude so as to reduce the effect of this non-linearity.

The output of the detector 64 is operatively coupled to a low pass filter 66, the output from which is operatively coupled to a DC offset removal circuit 68 which provides for adjustable DC offset removal, for example using an inverting amplifier that sums output from the low pass filter 66 with the output from a D/A converter under control of a microprocessor. This allows for subsequent DC amplification of the resulting signal. The output of the DC offset removal circuit 68 is then operatively coupled to an amplifier 70, which for example has two different outputs, each with a different associated gain, a low gain output with a gain of 4 and a high gain output with a gain of 20. In the exemplary system, the low pass filter 66 comprises a 2-pole low pass filter with a 1 KHz cut-off frequency. Moreover each of the output stages provides single pole low pass filtering with a cut-off frequency of 500 Hz.

The behavior of the in-phase and out-of-phase modes of operation of an elementary proximity sensor 10 is illustrated in Figs. 2-5 which provide the results two dimensional (2-D) electrostatic simulations--using a 2-D simulation package called Macsyma/PDEase --of the voltage fields around the proximity sensor 10 and the electric field intensity at the surface of the sense electrode 12, wherein the sense electrode 12 is 20 cm wide, the control electrodes 14 on each side thereof are each 10 cm wide and separated therefrom by a 3 cm gap, both the sense 12 and control 14 electrodes are located at Y=0 centered

about X=0, and +10 Volt is applied to the sense electrode 12. The simulation further comprised a grounded 15 cm diameter circular target (electric-field-influencing media 20) located 20 cm from the centerline of the sense electrode 12.

The proximity sensor 10 is simulated in the in-phase (IP) mode by applying a +10 Volt signal to the control electrodes 14. Referring to Fig. 2, the contours of the contour plot of the associated potential field of the proximity sensor 10 are largely spherical with the intensity falling off sharply close to the sensor. The field of the three electrodes is substantially identical to that of one large electrode, except that only the capacitance of the sense electrode 12 portion is actually measured. The boundary conditions of the target forces the voltage to be zero on the surface thereof, thereby increasing the gradient of the potential field, thereby increasing the electric field 16 and the charge on the sense electrode 12, thereby increasing the capacitance thereof.

Fig. 3 illustrates the electric field intensity at the surface of the sense electrode 12 for the in-phase (IP) mode illustrated in Fig. 2. Because the sense electrode 12 is a conductor, the E-field is perpendicular at the surface thereof, although the direction of the E-field changes as the E-field moves out into free space. Also, the charge distribution on the sense electrode 12 is proportional (by the dielectric constant) to the normal E-field intensity.

Accordingly, Fig. 3 is also a plot of the charge density on the sense electrode 12. The capacitance can be calculated by integrating the charge and dividing by the applied voltage (C=Q/V).

Alternately, the simulation provides the integral of the E-field distribution, from which the in-phase capacitance Clp of the sense electrode 12 can be calculated by: £oZlength |Enonnaldg46pFC =""" = 8. 854 9. 262. 2/10=1. v This assumes that the height of the sensor is 20 cm. The result is approximate because a real 3-D sensor would have edges, and only the capacitance contribution of the top side of the sensor is included.

For capacitive sensors, the charge distribution on the associated electrode is typically U-shaped as shown in Fig. 3, which results from the boundary condition that the tangential E-field be zero on the conductor. This smoothing effect makes it difficult to even theoretically get much information about the location and shape of possible targets from

the charge distribution. Basically, the charge on the sensor can be broken into two (or four for 3-D cases) edges and a center region. Measuring these three (or five) values could give information about the location of targets. Detailed analysis of the shape of the charge distribution would principally be useful for relatively close targets. The sensor essentially maps the entire complex target space onto this U-shaped distribution.

The proximity sensor 10 is simulated in the out-of-phase (OP) mode by applying a-10 Volt signal to the control electrodes 14. Fig. 4 illustrates a contour plot of the potential field for the same geometric configuration as in Fig. 2, but with the control electrodes 14 switched to the out-of-phase (OP) mode. Unlike the in-phase case, in which the contours are largely spherical, the out-of-phase case has a zero voltage interface in the gaps between the sense 12 and control 14 electrodes. The voltages for the out-of-phase (OP) mode cancel relatively close to the proximity sensor 10 so that the potential field does not extend relatively far therefrom, resulting in a relatively short range proximity sensor 10.

Because the potential falls off sharply, grounded targets (electric-field-influencing media 20) would need to be located relatively close to the proximity sensor 10 in order to have an effect on the associated electric field 16. In the configuration of Fig. 4, the target is barely disrupting the field and so that its effect on the capacitance of the sense electrode 12 is relatively small.

Referring to Fig. 5, the E-field and charge distribution for the out-of-phase (OP) mode has a similar U-shape as for the associated in-phase (IP) mode illustrated in Fig. 3, but with substantially higher magnitudes, particularly at the edges, because of the high field strengths between the sensor and control electrodes. For the out-of-phase (OP) mode, the maximum E-field at the edge is 2200 V/m, while for the in-phase (IP) mode, the corresponding maximum is only 125 V/m. The E-field strength at the center is about 100 V/m for the out-of-phase (OP) mode, compared with about 40 V/m for the in-phase (IP) mode.

The out-of-phase capacitance Cop of the sense electrode 12 is given by: OZlength f Enormal == 8. 854*36. 73*. 2/10=6. 5pF Accordingly, the out-of-phase capacitance Cop--about 6.5 pF--of the sense electrode 12 is about four times the in-phase capacitance Clp--about 1.6 pF. The

absolute capacitance of the sensor in the out-of-phase (OP) mode is larger than that of the in-phase (IP) mode, but the in-phase (IP) capacitance increases more sharply as targets become closer to the sense electrode 12.

Fig. 6 illustrates a comparison of the simulation results for the in-phase capacitance Clp and out-of-phase capacitance Cop as a function of the distance to a relatively large conductive sheet target. The out-of-phase capacitance Cop is larger, but is practically level after 4 inches. The in-phase capacitance Clp is much smaller, but almost equals the out-of-phase capacitance Cop for close-in targets. The difference Cop-Cjp (hereinafter referred to as"DeltaCap") decreases as the target becomes closer.

Accordingly, DeltaCap provides a measure of target distance, and by using this measure, this also provides a means for compensating for offsets that are common to both the in-phase (IP) and out-of-phase (OP) measurements, because switching the sensor from in-phase (IP) to out-of-phase (OP) modes principally affects the fields around the proximity sensor 10, and offsets in the associated first circuit 42 common to both measurements are cancelled during the calculation of DeltaCap.

Whereas the means measuring capacitance by the first circuit 42 provides a voltage that is roughly linear with the sensor capacitance, there are, however, a number of associated gain factors in the first circuit 42 are susceptible to drift. Referring to Figs. 7a- f, a switchable calibration capacitor CCAL can be used to compensate for this drift, and to provide a sensitivity factor in units of V/pF that can be used to calibrate the system.

Operatively coupling a relatively small (e. g. 1 pF or less) capacitance can be subject to error, depending upon the associated switching arrangement. For example, in Figs. 7a-b, one terminal of the calibration capacitor CCAL is connected to the first node 48, and the other terminal thereof is connected to ground through a switch S1. Referring to Fig. 7b, whereas this configuration is satisfactory when the switch Sl is closed, referring to Fig.

7a, when the switch is opened the capacitance of the switch S1 is typically larger than that of the calibration capacitor CCAL. For example, a typical FET may have an OFF capacitance of 40 pF, so that if the calibration capacitor CCAL is 1 pF, then the series combination is 0.98 pF, which means that the calibration capacitor CCAL is effectively never switched out of the circuit by the switch S1.

Referring to Figs. la and 7c-f, a first terminal 72 of a first switch S1 is operatively coupled to a first terminal 74 of a second switch S2 at a second node 76, and a first terminal 78 of a capacitor CCAL is operatively coupled to the second node 76. A second terminal 80 of the second switch S2 is operatively coupled to a first input 82 of an operational amplifier 84. Referring to Figs. 7c-d, a second terminal 86 of the capacitor CCAL is operatively coupled to a second input 88 of the operational amplifier 84 and a second terminal 90 of the first switch S1 is operatively coupled to a circuit ground 92.

Referring to Figs. 7e-f the second terminal 90 of the first switch S1 is operatively coupled to the second input 88 of the operational amplifier 84 and the second terminal 86 of the capacitor CCAL is operatively coupled to the circuit ground 92.

Referring to Figs. 7d and 7f, when the first switch S1 is closed and the second switch S2 is open, one side of capacitor CCAL is pulled to ground and the capacitor CCAL is operatively coupled to the second node 76, so as to add the capacitance CCAL thereto.

Referring to Figs. 7c and 7e, when the first switch S1 is opened and the second switch S2 is closed, one terminal of the capacitor CCAL is driven by a buffered version of the signal on the other terminal of the capacitor CCAL as a result of the property of the operational amplifier 84 to substantially null a potential difference between the associated first 82 and second 88 inputs thereof. To the extent that these signals have identical voltages, there is substantially no current flowing through the capacitor CCAL and the capacitor CCAL is effectively switched out.

Referring to Fig. 8a, illustrating an example of an algorithm (800) to detect the proximity of an object to a proximity sensor 10, in step (802) the second oscillatory signal 30 is switched to the out-of-phase (OP) mode by the control electrode driver 32 and the control electrodes 14.1,14.2 and 14.3 are driven with an out-of-phase (OP) signal. In step (804) the associated third signal 50 (Vop) is measured by the first circuit 42. Then in step (806), the calibration capacitor CCAL is operatively coupled to the first node 48 by activating FET transistor Qla and deactivating FET transistor Q2a, and in step (808) the associated fifth signal 94 (VCAL) is measured by the first circuit 42, after which in step (810) the calibration capacitor CCAL is decoupled from the first node 48 by deactivating FET transistor Qla and activating FET transistor Q2a. In step (812), the bias capacitor CBIAS is operatively coupled the first node 48 by activating FET transistor Qlb and deactivating FET transistor Q2b, and in step (814) the second

oscillatory signal 30 is switched to the in-phase (IP) mode by the control electrode driver 32 and the control electrodes 14.1,14.2 and 14.3 are driven with an in-phase (IP) signal.

In step (816) the associated fourth signal 52 (Vlr) is measured by the first circuit 42, after which in step (818), the bias capacitor CBIAS is decoupled from the first node 48 by deactivating FET transistor Qlb and activating FET transistor Q2b.

In step (820), a sensitivity factor for calibrating the first circuit 42 is calculated, in units of Volts per unit of capacitance, for example V/pF, as follows: Sensitivity = (VcAL-Vop)/CcAL In step (822), the DeltaCap measure is calculated using this sensitivity factor, a follows: DeltaCap = (Vop-Vip)/Sensitivity + CBIAS When no objects are present, the DeltaCap measure is maintained within a range by the above described calibration process. Referring to Fig. 9, illustrating a one-dimensional plot of the DeltaCap measure, should the DeltaCap measure drop below the normal minimum, into the"always disable"region, this would indicate the presence of a target in the sensing region. For example, with the proximity sensor 10 used to detect an object proximate to a restraint actuator 96 of a safety restraint system, this would indicate that the restraint actuator 96, for example and air bag inflator, should be disabled so at to prevent injury to an occupant proximate thereto as a result of the deployment thereof.

When the DeltaCap measure is in the normal region, a long term average of the DeltaCap measure is calculated, and is referred to herein as the offset. If the the DeltaCap measure drifts slowly due to temperature or time, the offset slowly tracks these changes. If the DeltaCap measure drops quickly by a given threshold amount below the offset, this would indicate the presence of a target proximate to the proximity sensor 10 so that the restraint actuator 96 would be disabled. However, over time the offset would adjust down and the restraint actuator 96 would become re-enabled. The DeltaCap measure is allowed to drift within the normal region and is slowly tracked by the offset. For small changes in the DeltaCap measure (for example as a result of far targets), a threshold below the present offset is used for disabling the restraint actuator 96. As a target moves close to the proximity sensor 10 and the DeltaCap measure decreases significantly, the DeltaCap

measure enters the"always disable"region. When the DeltaCap measure is in the"always disable"region, the offset is not updated.

A special condition is when a large dielectric object is placed on the proximity sensor 10, or if the proximity sensor 10 is sprayed with water, wheren the DeltaCap measure may be shifted up, which is opposite to the change caused by a grounded target. If the DeltaCap measure is above the normal maximum, then special techniques are necessary for detecting targets, as described hereinbelow.

Returning to Fig. 8a, in step (824), if the DeltaCap measure exceeds the normal maximum, then in step (850) a dielectric processing algorithm is called as will be described hereinbelow in conjunction with Fig. 8b. Otherwise, if in step (826) the DeltaCap measure is less than the normal minimum, then the restraint actuator 96 is disabled in step (832) and the process repeats with step (802). Otherwise from step (826) the offset, for example a running average of the DeltaCap measure, is updated in step (828) and if in step (830) the offset exceeds the DeltaCap measure by more than a threhsold, the restraint actuator 96 is disabled in step (832) and the process repeats with step (802).

Otherwise, from step (830), the process repeats with step (802).

The DeltaCap measure provides a difference between an out-of-phase (OP) capacitance and an in-phase (IP) capacitance, which inherently reduce drifts because any common-mode effects in the first circuit 42 are substantially cancelled by the differencing process. However, one undesirable side-effect of the DeltaCap measure is an is an increased sensitivity to dielectric objects. Referring to Figs. l0a-b, the out-of-phase (OP) capacitance of Fig. 10a is increased more than the in-phase (IP) capacitance of Fig. 10b for dielectric objects close to the sensor. The capacitance is increased by the dielectric object in proportion to the amount of the electric field in the dielectric and the dielectric constant of the object. In the out-of-phase (OP) mode, the fields are contained close to the electrode surface and so more of the electric field travels through the dielectric object. In the in- phase (IP) mode, the fields pass through the dielectric, but most of the field is in air, resulting in only a marginal increase of capacitance.

The principal manifestation of this effect is from thick dielectric objects relatively close to the proximity sensor 10, such as from books or from water on the proximity sensor 10.

Typically, one section of newspaper is not enough to increase the DeltaCap measure. A

dielectric object tends to increase the DeltaCap measure, which is the opposite to the effect of a person who is typically grounded, so the sensor would not normally cause the restraint actuator 96 to disable as a result of the presence of such an object.

This problem is mitigated by either 1) using an in-phase guard band around the sensor, 2) using the change in the in-phase (IP) capacitance as a back up measure for disabling the restraint actuator 96, or 3) using the changes in the in-phase (IP) and out-of-phase (OP) capacitances to deduce that a dielectric object is present and then rapidly updating the offset.

Referring to Fig. la, a proximity sensor 10 is illustrated with three control electrodes 14.1,14.2 and 14.3 and a guard 98 around the sense electrode 12. This is just and example one possible arrangement. The arrangement of the electrodes can generally be adapted in accordance with constraints of the region 18 to be sensed. As was illustrated in Fig. 5, the charge distribution for the out-of-phase (OP) mode has relatively high peaks at the edges. When the dielectric increases the capacitance of the sense electrode 12, the bulk of the extra charge travels to the edges of the sense electrode 12. The guard 98 is driven in-phase (IP) with the sensor, causing this extra charge to be on the guard, but because the guard is not electrically connected to the sense electrode 12, this charge is not measured. Accordingly, the guard reduces the effect of dielectrics on the out-of-phase (OP) capacitance of the sense electrode 12.

Fig. 11 illustrates the results of a simulation in which a 5 mm dielectric having a relative permittivity of 3 is placed at various distances. The guard reduces both the out-of- phase (OP) capacitance and the sensitivity to dielectrics. The effect of the dielectric is noticeable primarily within a short range, for example less than 3 inches.

Fig. 12 illustrates actual test data in which various dielectric objects are placed on a sensor, with and without the guard. The guard reduces the effect of these objects, except for the wet towel cases. Wet towels are particularly difficult because they are conductive as well as having a large dielectric constant.

A second technique of mitigating dielectric effects is to consider only the in-phase (IP) capacitance, since the in-phase (IP) capacitance is not drastically affected by dielectrics. When a target approaches the sensor, the in-phase (IP) capacitance increases.

Although the absolute value of the in-phase (IP) capacitance is not reliable, the change therein over time can be calculated.

A measure referred to as DeltaIP is calculated and used to calculate a measure called DynamicIP which this tracks the change of the in-phase capacitance, as follows: DeltalP (k) = Clp (k)-Clp (k-1) DynamiclP (k) = DynamiclP (k-1) + DeltalP (k)-dampingfactor (DynamiclP must be >= 0) As an example, if the DeltaCap measure is increased above the offset, the system is still be able to disable for a target moving into the danger zone by checking if the DynamicIP measure is greater than a threshold. This works even if a large book is over the sensor, but principally works for moving targets.

The third technique for mitigating the effect of dielectrics and water is to monitor the changes in the in-phase (IP) and out-of-phase (OP) capacitances and use that information to categorize the situation. Then, if necessary, the offset can be quickly updated. For example, grounded targets increase the in-phase (IP) capacitance more than the out-of- phase (OP) capacitance, while dielectric objects increase the out-of-phase (OP) capacitance more than in-phase (IP) capacitance.

Fig. 13 illustrates a grounded target moving closer to and farther away from the proximity sensor 10. Fig. 14 illustrates a dielectric (i. e. a magazine) moving away from the proximity sensor 10. For both a grounded target moving towards the sensor and for a dielectric object moving away from the sensor, the DeltaCap measure decreases, But for the grounded target case, most of the change is due to the in-phase (IP) capacitance, while for the dielectric case the out-of-phase (OP) change is greater than the in-phase (IP) change. If these cases can be separated, then the offset can be quickly updated for the dielectric case and slowly updated for the target case.

Referring to Fig. 8b, illustrating a dielectric processing algorithm (850), in step (852) an in-phase capacitance CIp is calculated from the fourth signal 52 (Vlp) and the sensitivity factor as follows: Cp = Vp/Sensitivity-CstAs

In step (854) a DeltaIP measure is calculated from the change in in-phase capacitance Cjp over time, as follows: DeltalP (k) = Clp (k)-Clp (k-1) In step (856) a DynamicIP measure is calculated using a damping factor, as follows: DynamiclP (k) = DynamiclP (k-1) + DeltalP (k)-dampingfactor If in step (858) the DynamicIP measure is less than zero, then in step (860) the present value of the DynamicIP measure is set to zero. Otherwise, if in step (862) the DynamicIP measure is greater than a threshold, then in step (864) the restraint actuator 96 is disabled and in step (866) the process returns to step (802). Otherwise from step (862), in step (868) a SumDeltaIP measure is calculated as a running sum of the DeltaIP measures over time. In step (870), an out-of-phase capacitance Cop is calculated from the third signal 50 (Vop) and the sensitivity factor as follows: CoP = Vop/Sensitivity In step (872) a DeltaOP measure is calculated from the change in out-of-phase capacitance Cop over time, as follows: DeltaOP (k) = Cop (k)-Cop (k-1) In step (874) a SumDeltaOP measure is calculated as a running sum of the DeltaOP measures over time. If in step (876) the SumDeltaIP measure is greater than the SumDeltaOP measure, then in step (878) a grounded target is assumed to be present, and in step (880) the offset is continued to be updated slowly, after which in step (866) the process returns to step (802). Otherwise from step (876), if in step (882) the SumDeltaOP measure is greater than the SumDeltaIP measure, then in step (884) a dielectric object is assumed to be present, and in step (886) the offset is updated quickly, after which in step (866) the process returns to step (802).

Fig. 15 illustrates a schematic of the front end amplifier 54, comprising an amplifier U, and a offset canceller U2, and the associated calibration and bias capacitor circuitry.

The amplifier U, can be understood as a current to voltage converter. The non- inverting input is driven with a sinusoid and because of the properties of the op-amp the same voltage is present on the inverting input. The signal on the inverting input drives the

sensor electrode and a AC current flows out of the sensor. This same current flows through the feedback resistor Rl and this generates the change in the output.

The circuit can also be understood as a non-inverting amplifier with a transfer function of :<BR> Vout = V+ (l + jcoRC) This circuit has the following properties: 1. If there is no capacitance then the output is the same as the drive signal on the non- inverting input (i. e. it is a voltage follower).

2. As the capacitance of the sensor increases, the output voltage increases and also starts to go out of phase with the drive voltage.

3. The change in output voltage is linearly related to the frequency of the drive signal, the size of the feedback resistor Ri, and the magnitude of the drive voltage, and the change in the capacitance.

4. If there is a large offset capacitance, then the gain must be low or the output voltage will be saturated. Accordingly, an offset canceling circuit is incorporated to prevent this problem.

5. The circuit stops working well above about 100KHz because the op-amps begin to become non-ideal. The circuit requires that the feedback current is such that the inverting input is kept at the same voltage as the non-inverting input.

The following design considerations can be used in configuring the first circuit 42: 1. Higher operating frequencies provide for higher associated current to the sense electrode 12, but the frequency is preferably limited to that range for which the performance of the operational amplifier remains reasonably ideal.

2. The size of the drive signal is preferably as large as possible for increased signal- to-noise ratio, but is preferably not so large as to saturate or rail output voltage the measure of capacitance increases. Generally, both noise currents and signal currents are amplified, so it is desireable to make the signal currents as large as possible.

3. The feedback resistor R, can be set to provide the desired sensitivity. The change in output voltage due a change in capacitance is: AVout (dc) = a x m x Ri x Vdrive (peak) x Gain (dc) x ACapacitance where a is dependent on the circuit capacitance but is usually 0.7-0.9. For example, if it is desired to have a 40mV output change for a 0.0lpf change in capacitance then with a = 0.8, m = 628 x 10^5, Vdrive (peak) = IV, Gain (dc) = 20, ACapacitance = 10^-14 R, =. 04/ (0.8 x 6.28^5 x 1 x 20 x 10^-14) = 398KQ The remainder of the circuit is used to adjust the current going in and out of the inverting input node. A large offset capacitance can cause a large offset current which can saturate the output voltage at the desired gain level. These offset capacitances can come from the circuit, or the wire going to the sensor, or the back side of the sensor. Amplifier U2 is designed to generate a signal which is in-phase with the sensor drive signal (inverting input). If the output of U2 is made larger, it will inject current that cancels some of this offset current. Buffer amplifiers U3 and U4 allow the drive to amplifier U2 and to amplifier Ul to be adjusted in amplitude while staying in phase.

It will be understood by one of ordinary skill in the art that the means for controlling the state of the second oscillatory signal, the means for measuring the response signal from first electrode, the means for means for forming a difference of the third and fourth signals, the means for calibrating the difference, and the means for generating a measure of proximity of an object to the proximity sensor can be accomplished by various analog or digital circuits or by software using a computer, for example a microprocessor.

It should be understood that the first oscillatory signal 26 in general need not be either sinusoidal or periodic. Generally, an in-phase signal corresponding to the second state 40 undergoes similar transitions to the first oscillatory signal 26, and an out-of-phase signal corresponding to the first state 38 undergoes substantially opposite transitions to that of the first oscillatory signal 26. Whereas the signal generator 28 is illustrated herein as a sinusoidal oscillator, other types of signal generators 28 can be used, for example a logic circuit. The DC bias to the second oscillatory signal 30 can be either the same or different for different states 36.

The proximity sensor 10 can be used in a variety of applications, for example in a vehicle for detecting the proximity of an occupant to an air bag inflator so that the air bag inflator can be disabled if the occupant becomes located within an at-risk region proximate to the air bag inflator. For example, the proximity sensor can be located in a seat, in the instrument panel, in a cover to the air bag inflator, or in the steering wheel. The proximity sensor 10 may also be used for other proximity sensing applications.

While specific embodiments have been described in detail, those with ordinary skill in the art will appreciate that various modifications and alternatives to those details could be developed in light of the overall teachings of the disclosure. Accordingly, the particular arrangements disclosed are meant to be illustrative only and not limiting as to the scope of the invention, which is to be given the full breadth of the appended claims and any and all equivalents thereof.