Split electrode to minimize charge transients, motor amplitude mismatch errors, and sensitivity to vertical translation in tuning fork gyros and other devices

Information

  • Patent Grant
  • 5911156
  • Patent Number
    5,911,156
  • Date Filed
    Monday, February 24, 1997
    27 years ago
  • Date Issued
    Tuesday, June 8, 1999
    25 years ago
Abstract
A micromechanical tuning fork gyroscope having two center electrodes is disclosed. The two center electrodes are excited with bias potentials of opposite polarity. The oppositely biased center electrodes provide electrical symmetry across the gyroscope and thereby reduce charge transients and sensitivity to vertical translation. Currents injected directly into the proof masses are equal and opposite and thus cancel. Motor lift forces acting on the proof masses and interleaved electrodes are equal, and hence the proof masses move in pure translation, thereby reducing in-phase bias. Further, any pure translation normal to the plane of the gyroscope does effect sense axis output signals.
Description

CROSS-REFERENCE TO RELATED APPLICATIONS
Not Applicable
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT
Not Applicable
BACKGROUND OF THE INVENTION
Micromechanical tuning fork gyroscopes such as that shown in FIG. 1 are known. The tuning fork gyroscope includes silicon proof masses which are suspended by support flexures above a glass substrate, and comb electrodes which are used for oscillating the proof masses. Metallic sense electrodes are disposed on the glass substrate below the proof masses for detecting Coriolis motion by indicating out-of-plane capacitance changes. Because the tuning fork gyroscope operates in this manner, it is desirable that the amplitude of the oscillation be held at a predetermined constant in order to provide a more accurate output indicating rate.
The amplitude of the oscillating motor of tuning fork gyroscopes is typically controlled by a conventional servo loop connected to a single capacitive in-plane pick-off ("center electrode"). In this technique motor position is converted to a proportional voltage by measuring charge variation on the center electrode which is biased with a DC voltage. The resulting motor position signal is amplified and detected by a full wave rectifier. The rectifier output is then filtered, and the filtered voltage compared against a reference voltage; the difference forming an error voltage. This error voltage is then used to control the motor drive amplitude using a loop controller to adjust motor amplitude to a predetermined constant. However, this particular technique has a potential drawback.
The conventional servo loop technique may have some instability in the center electrode. The sensitivity of the DC biased center electrode varies slowly over time due to a spurious charge accumulation on the glass substrate beneath the center electrode. As this charge accumulates on the glass, the sensitivity of the center electrode is modified. In response, the loop controller calls for modified drive force to compensate. The result is a motor amplitude transient wherein amplitude changes over time as substrate charge accumulates. This will result in lower accuracy than is otherwise possible because of the relationship between amplitude and Coriolis force in the system.
BRIEF SUMMARY OF THE INVENTION
In accordance with the present invention, a tuning fork gyroscope includes a plurality of center and outer electrodes. The total proof mass structure comprises two independent masses, a right and a left, connected by a series of beams and flexures. Charge into the proof mass structure is the mechanism by which Coriolis Force is measured. Center and outer motor plurality allows generation and detection of proof mass motion while minimizing charge injection into the total proof mass structure resulting from inequalities in the amplitude and/or phase of the relative proof masses. By forming an electrode configuration that provides each of the masses with split center and outer motor electrodes, errors from mismatch in relative amplitude or phase can be rejected. By applying excitations of equal and opposite potentials to each set of independent center and outer motor electrodes, each mass cancels the charge generated by its own motion, thereby reducing the in-phase bias errors and minimizing limitations on dynamic range. Because each proof mass interacts with independent split center and outer electrodes having equal and opposite potentials, the net charge generated in the total proof mass structure from amplitude mismatch errors is minimized.
Splitting both the center and outer motor electrodes desensitizes the gyro to errors from charge injection into the proof mass from mismatch in the amplitude between the right and left proof masses. Charge injection results from each proof mass' interaction with the excitations used to provide both the electrostatic forcing and detection of proof mass motion. A net charge injection occurs if the charge generated from the right proof mass and left proof mass are not equal and opposite, a prevailing condition if the oscillatory displacements of the right and left proof masses are mismatched in amplitude and/or phase. By splitting the center and outer motor electrodes evenly and applying excitations of opposing magnitude, each proof mass cancels the charge inherent in its own motion, thereby reducing in-phase bias errors and dynamic range limitations.
The center electrodes reduce substrate charging effects and reduce undesirable motor lift forces by providing equal numbers of oppositely biased center electrodes. The center electrodes are arranged to provide electrical symmetry across the tuning fork gyroscope. Because of the symmetry, voltages induced in the substrate by the center electrodes are equal and opposite so that the effect of substrate charging on in-phase bias is reduced. Further, currents injected directly into the proof masses are equal and opposite, and therefore tend to cancel. As a result, motor lift forces are equal and the proof masses move in pure translation, thereby reducing in-phase bias. The net current injected into the proof mass is the gyroscope output signal. This current flows through the anchors into a transimpedence amplifier which converts charge (the integral of current) into an output voltage. The transimpedence amplifier holds the proof mass at virtual ground. Maintaining electrical symmetry greatly reduces erroneous signals from in-plane motion, from common mode proof mass translation normal to the substrate, and from charge transients. With opposite biases on the sense electrodes, the desired gyroscope output is the differential vertical displacement.





BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWING
The invention would be more fully understood in light of the following detailed description of the drawing in which:
FIG. 1 is a diagram of a prior art tuning fork gyroscope;
FIG. 2 is a diagram of a tuning fork gyroscope with a plurality of center electrodes;
FIG. 3 is an alternative configuration of the tuning fork gyroscope of FIG. 2;
FIGS. 4 and 5 illustrate circuits for applying motor bias; and
FIG. 6 is a diagram of the tuning fork gyroscope with a plurality of outer motor electrodes.





DETAILED DESCRIPTION OF THE INVENTION
A micromechanical tuning fork gyroscope is illustrated in FIG. 2. The tuning fork gyroscope includes first and second proof masses 3a, 3b, first and second motor electrodes 5a, 5b, first and second sense electrodes 7a, 7b, first and second center electrodes 9a, 9b, and a substrate 11. The center electrodes, sense electrodes and motor electrodes are disposed on the substrate. The proof masses are disposed over the sense electrodes, being supported by flexures 13a, 13b. The flexures are attached to the substrate at anchor points 15 and permit movement of the proof masses relative to the sense electrodes. Each proof mass includes combs extending outward from first and second sides thereof. The center electrodes and motor electrodes also include combs. The combs of motor electrode 5a are interleaved with the combs of proof mass 3a, the combs of center electrode 9a are interleaved with the combs of proof mass 3a, the combs of center electrode 9b are interleaved with the combs of proof mass 3b and the combs of motor electrode 5b are interleaved with the combs of proof mass 3b.
The operation of the tuning fork gyroscope is electromechanical. Time varying drive signals 17a, 17b are provided to motor electrodes 5a, 5b, respectively. The drive signals generate electrostatic coupling between interleaved combs 19a, 19b, 21a, 21b, attached to the motor electrodes 5a, 5b and proof masses 3a, 3b, respectively, and impart an oscillating force upon the proof masses along a motor drive axis 23. The oscillating force causes the proof masses to oscillate in a plane of vibration 25. In response to an inertial input such as a rotational rate, the proof masses deflect out of the plane of vibration. Sense biases +V.sub.s, -V.sub.s are applied to the sense electrodes 7a, 7b, respectively to establish a potential between the sense electrodes 7a, 7b and the proof masses 3a, 3b, respectively, so that changes in capacitance between the electrodes and the adjacent proof masses as a result of the deflection out of the plane of vibration can be measured.
Measurement of an inertial input with the tuning fork gyroscope is based upon the principal of Coriolis force.
F.sub.c =2m.OMEGA.XV (1)
Where
m is mass
V is the proof mass velocity and
.OMEGA. is the input rate.
Mass and velocity are known for the tuning fork gyroscope. Hence, inertial input motion can be measured based upon charge variation between the proof masses and sense electrodes. However, in order to achieve accurate results it is important that the proof mass velocity remain constant.
An oscillator circuit 27 is employed to measure proof mass velocity from at least one of the center electrodes 9a, 9b and, in response thereto, vary the drive signals 17a, 17b to compensate for variations in velocity. Bias potentials +V.sub.B, -V.sub.B are applied to the center electrodes 9a, 9b, respectively, in order to facilitate measurement of proof mass velocity through feedback signals 29a, 29b. The bias signals +V.sub.B, -V.sub.B are coupled to the center electrodes 9a, 9b through resistors, 31a, 31b. Charge variations caused by displacement of the proof masses in the plane of vibration are then detected and employed as feedback. The bias signals +V.sub.B, -V.sub.B may be DC voltage, AC voltage or a combination AC+DC voltage. Further, the bias signals are equal in magnitude and opposite in polarity. Circuits for applying the motor bias are shown in FIGS. 4 and 5. In FIG. 4, the bias may only be DC, but in FIG. 5 AC or DC or AC+DC is suitable.
The changing proximity between the proof mass and the adjacent center electrode, which results in charge variations, is indicated through the electrostatic coupling of the interleaved combs. As the proof mass oscillates, proximity changes over time. Consequently, the potential between the interleaved comb electrodes changes over time. The rate of change of the potential of the feedback signals from the center electrodes are thus indicative of proof mass velocity. In order to maintain constant proof mass velocity, the feedback signals are compared. with reference signals and the result of the comparison is employed to adjust the drive signals.
The oppositely biased center electrodes reduce the effect of undesirable substrate charging by providing electrical symmetry between left and right sides of the tuning fork gyroscope. Symmetry exists where for each bias applied to the tuning fork gyroscope another bias of equal magnitude and opposite polarity exists and where the gyroscope can be bisected into two regions of equal and opposite electrical characteristics. Symmetry reduces the effects of charge transients and sensitivity to vertical translation because the oppositely biased signals applied to the center electrodes tend to cancel. For example, voltages induced in the gyroscope substrate by the bias potentials are equal and opposite, so that substrate charging effect on in-phase bias are reduced. Further, motor lift forces acting on the proof. masses and interleaved comb electrodes are equal and hence the proof masses move in pure translation, thus reducing in-phase bias. Another benefit of symmetry is that pure translation normal to the plane of the gyroscope does not produce a sense axis output. Thus, the sense electrode output only reflects actual inertial motion. The net current injected into the proof mass is the gyroscope output signal. This current flows through the anchors into a transimpedence amplifier which converts charge (the integral of current) into an output voltage. The transimpedence amplifier holds the proof mass at virtual ground. Maintaining electrical symmetry greatly reduces erroneous signals from in-plane motion, from common mode proof mass translation normal to the substrate, and from charge transients. With opposite biases on the sense electrodes, the desired gyroscope output is the differential vertical displacement. For these reasons the center electrodes are disposed symmetrically upon the substrate.
FIG. 3 illustrates an alternative center electrode configuration. In the alternative embodiment, the center electrodes 9a, 9b each include first and second respective sets of comb electrodes 33a, 33b, 35a, 35b, which are interleaved with the comb electrodes 37, 39 of the proof masses 3a, 3b, respectively. That is, each center electrode interacts with both proof masses. As with the previously described embodiment, the center electrodes have bias potentials +V.sub.B, -V.sub.B applied thereto, respectively, in order to facilitate measurement of proof mass velocity through feedback signals 41, 43. The bias potentials may be DC voltage, AC voltage or a combination DC+AC voltage. Because each center electrode provides a measurement of the velocity of both proof masses, a single feedback signal from one of the center electrodes may by utilized by the oscillator circuit to maintain a constant proof mass velocity. Alternatively, a differential readout 45 may be employed with the feedback signals from each center electrode to provide an indication of proof mass velocity. Since each of the center electrodes interact with both the proof masses, currents injected into the proof masses through the center electrodes are equal and opposite, and thus effectively cancel.
Another alternative embodiment is illustrated in FIG. 6. In this embodiment the center electrodes 9a, 9b are split as described with respect to FIG. 3, above. Additionally, the gyro includes split left 38a, 38b and right 38c, 38d motor electrodes. To achieve symmetry +VAC is applied to electrodes 38a, 38c, and -VAC is applied to electrodes 38b, 38d.
It will now be apparent in view of the above description that the present invention defines a method for sensing oscillatory motion of an oscillating mass. Sensing oscillatory motion includes providing an even number of oscillatory motion sensing elements, biasing first and second groups of the oscillatory motion sensing elements with first and second bias potentials of opposite polarity, the first and second groups of oscillatory motion sensing elements being equal in number, and then sensing oscillatory motion with at least one of the oscillatory motion sensing elements. By arranging the oscillatory motion sensing element in equal groups which are biased with signals of opposite polarity, stray current and voltage injection into other elements of the device tend to cancel. Such injection is typically through interleaved comb electrodes, and arrangement of oscillatory motion sensing elements can be such that each element is coupled to only one oscillating mass, or such that each element is coupled to more than one oscillating mass. Depending on the arrangement, the symmetry of the device will change as described above. Hence, the technique of balancing the injected current through an even number of electrodes also applies to rotary vibrating gyroscopes.
It should be understood that various changes or modifications may be made from the embodiment being disclosed. Accordingly, the invention is not to be viewed as limited except by the scope and spirit of the appended claims.
Claims
  • 1. A micromechanical tuning fork gyroscope for measuring an inertial input, comprising:
  • a substrate;
  • first and second sense electrodes disposed symmetrically on said substrate;
  • at least first and second proof masses disposed over said first and second sense electrodes, respectively, said first and second proof masses having comb electrodes extended from respective inner and outer sides thereof;
  • first and second motor electrodes disposed symmetrically on said substrate, said first and second motor electrodes having comb electrodes, said comb electrodes of said first and second motor electrodes interleaved with the outer side comb electrodes of said first and second proof masses, respectively; and first and second oppositely biased center electrodes disposed symmetrically on said substrate, said first and second center electrodes having comb electrodes at least one of said center electrodes comb electrodes being interleaved with innerside comb electrodes of said first proof mass and providing a feedback signal indicative of velocity of said first proof mass.
  • 2. The tuning fork gyroscope of claim 1 wherein said comb electrodes of said second center electrode are interleaved with said innerside comb electrodes of said second proof mass.
  • 3. The tuning fork gyroscope of claim 2 wherein said first portion center electrode includes comb electrodes interleaved with a portion of said comb electrodes of said second proof mass and said second center electrode includes comb electrodes interleaved with a portion of said comb electrodes of said first proof mass.
US Referenced Citations (116)
Number Name Date Kind
RE32931 Staudte May 1989
RE33479 Juptner et al. Dec 1990
3053095 Koril et al. Sep 1962
3251231 Hunt et al. May 1966
3370458 Dillon Feb 1968
3696429 Tressa Oct 1972
3913035 Havens Oct 1975
4044305 Oberbeck Aug 1977
4122448 Martin Oct 1978
4144764 Hartzell, Jr. Mar 1979
4155257 Wittke May 1979
4234666 Gursky Nov 1980
4321500 Paros et al. Mar 1982
4336718 Washburn Jun 1982
4342227 Petersen et al. Aug 1982
4381672 O'Connor et al. May 1983
4406992 Kurtz et al. Sep 1983
4411741 Janata Oct 1983
4414852 McNeill Nov 1983
4447753 Ochiai May 1984
4468584 Nakamura et al. Aug 1984
4478076 Bohrer Oct 1984
4478077 Bohrer et al. Oct 1984
4483194 Rudolf Nov 1984
4484382 Kawashima Nov 1984
4490772 Blickstein Dec 1984
4495499 Richardson Jan 1985
4499778 Westhaver et al. Feb 1985
4502042 Wuhrl et al. Feb 1985
4522072 Sulouff et al. Jun 1985
4524619 Staudte Jun 1985
4538461 Juptner et al. Sep 1985
4585083 Nishiguchi Apr 1986
4590801 Merhav May 1986
4592242 Kempas Jun 1986
4596158 Strugach Jun 1986
4598585 Boxenhorn Jul 1986
4600934 Aine et al. Jul 1986
4619001 Kane Oct 1986
4621925 Masuda et al. Nov 1986
4628283 Reynolds Dec 1986
4629957 Walters et al. Dec 1986
4639690 Lewis Jan 1987
4644793 Church Feb 1987
4651564 Johnson et al. Mar 1987
4653326 Danel et al. Mar 1987
4654663 Alsenz et al. Mar 1987
4665605 Kempas May 1987
4670092 Motamedi Jun 1987
4671112 Kimura et al. Jun 1987
4674180 Zavracky et al. Jun 1987
4674319 Muller et al. Jun 1987
4679434 Stewart Jul 1987
4680606 Knutti et al. Jul 1987
4699006 Boxenhorn Oct 1987
4705659 Bernstein et al. Nov 1987
4706374 Murkami Nov 1987
4712439 North Dec 1987
4727752 Peters Mar 1988
4735506 Pavlath Apr 1988
4736629 Cole Apr 1988
4743789 Puskas May 1988
4744248 Stewart May 1988
4744249 Stewart May 1988
4747312 Herzl May 1988
4750364 Kawamura et al. Jun 1988
4761743 Wittke Aug 1988
4764244 Chitty et al. Aug 1988
4776924 Delapierre Oct 1988
4783237 Aine et al. Nov 1988
4789803 Jacobsen et al. Dec 1988
4792676 Hojo et al. Dec 1988
4805456 Howe et al. Feb 1989
4808948 Patel et al. Feb 1989
4815472 Wise et al. Mar 1989
4834538 Heeks et al. May 1989
4851080 Howe et al. Jul 1989
4855544 Glenn Aug 1989
4869107 Murakami Sep 1989
4881410 Wise et al. Nov 1989
4882933 Petersen et al. Nov 1989
4884446 Ljung Dec 1989
4890812 Chechile et al. Jan 1990
4893509 MacIver et al. Jan 1990
4898032 Voles Feb 1990
4899587 Staudte Feb 1990
4900971 Kawashima Feb 1990
4901586 Blake et al. Feb 1990
4916520 Kurashima Apr 1990
4922756 Henrion May 1990
4929860 Hulsing, II et al. May 1990
4981359 Tazartes et al. Jan 1991
5001383 Kawashima Mar 1991
5013396 Wise et al. May 1991
5016072 Greiff May 1991
5025346 Tang et al. Jun 1991
5038613 Takenaka et al. Aug 1991
5055838 Wise et al. Oct 1991
5060039 Weinberg et al. Oct 1991
5090809 Ferrar Feb 1992
5094537 Karpinski, Jr. Mar 1992
5138883 Paquet et al. Aug 1992
5195371 Grieff Mar 1993
5203208 Bernstein Apr 1993
5205171 O'Brien et al. Apr 1993
5216490 Greiff et al. Jun 1993
5226321 Varnham et al. Jul 1993
5233874 Putty et al. Aug 1993
5241861 Hulsing, II Sep 1993
5349855 Bernstein et al. Sep 1994
5576250 Diem et al. Nov 1996
5604312 Lutz Feb 1997
5635638 Geen Jun 1997
5728936 Lutz Mar 1998
5757103 Lee et al. May 1998
5780739 Kang et al. Jul 1998
Foreign Referenced Citations (1)
Number Date Country
55-121728 Sep 1980 JPX