Note: Descriptions are shown in the official language in which they were submitted.
CA 02213245 1997-08-18
WO 96/28735 1'CT/L1S95/03011
1
MECI~ANICAL RESONANCE SILICON ACCELEROMETER.
T~HNICAL FIELD OF THE INVENTION
The present invention relates to accelerometers and, more particularly, to
vibrating silicon beam sensors arranged to sense acceleration.
BACKGROUND OF T E INVENTION
An accelerometer is a device which senses acceleration, as v~~ell as shocks
and
vibrations, along or about an input or sensitive axis. One type of such an
accelerometer
is a mechanical resonating accelerometer which senses linear accelerations
parallel to, or
along, an input axis. If acceleration is to be sensed three-dimension;~lly, a
triad of such
accelerometers is arranged such that a first accelerometer senses acceleration
along the x
coordinate axis, a second accelerometer senses acceleration along the y
coordinate axis,
and a third accelerometer senses acceleration along the z coordinate axis.
A linear accelerometer typically includes a damped seismic mass which is
positionally constrained by spring forces. In response to an acceleration, the
seismic
1 S mass moves relative to its support and, when the acceleration ends, the
seismic mass is
returned to its initial position by the spring forces. The displacement of the
seismic
mass due to acceleration is converted into an electrical output by various
types of
transducers in order to produce a measure of the acceleration.
For example, in a potentiometric accelerometer, the transducer is a
potentiometer
having a resistance held in a fixed position with respect to a support
surface. A wiper
arm of the potentiometer is driven by a mechanical linkage connected between a
seismic
mass and the wiper arm. As the seismic mass moves in response to
accelerations, the
mechanical linkage moves the wiper arm over the resistance of the
potentiometer to
change the electrical output from the potentiometer. This change in electrical
output
provides an indication of the amount and direction of acceleration.
An inductive type accelerometer typically uses an inductancf: bridge sensitive
to
the motion of a seismic mass. As the seismic mass moves in response to
accelerations,
a
the seismic mass drives a ferromagnetic armature with respect to two inductive
coils
resulting in an increase of the inductance of one inductive coil and a
decrease of the
inductance of the other inductive coil. The difference in inductances between
the two
inductive coils provides an indication of the amount and direction oi'
acceleration.
CA 02213245 1997-08-18
WO 96/28735 PCT/US95/03011
2
A strain gauge accelerometer includes a seismic mass attached to a strain gage
which may be fabricated out of metal wire, metal foil or semiconductors. Servo
accel-
erometers and piezoelectric accelerometers are also known. In piezoelectric
accelerometers, a seismic mass is mechanically connected to a crystal material
which
may be comprised of quartz or of such ceramic mixtures as titanate, niobate,
or
zirconate.
A typical prior art mechanical resonating linear accelerometer generally
utilizes
at least one, and more often two, quartz beams and is a rather complex
mechanical
assembly. In such a quartz beam linear accelerometer, a quartz beam is caused
to
vibrate at a base frequency. The quartz beam converts its mechanical vibration
into an
electrical signal which has a frequency which tracks the frequency of the
mechanical
vibration. In the presence of acceleration, the vibration frequency of the
quartz beam
changes and this change in vibration frequency provides an indication of the
amount and
direction of acceleration experienced by the quartz beam accelerometer.
If the quartz beam linear accelerometer employs two quartz beams, the two
quartz beams are generally arranged so that, in the presence of an
acceleration along an
input (i.e. sensitive) axis, one of the quartz beams experiences an increase
in vibration
frequency and the other quartz beam experiences a decrease in vibration
frequency. The
difference between these vibration frequencies of the two quartz beams
provides an
indication of the amount and direction of acceleration along the input axis.
Quartz
beams which are arranged in this push/pull manner benefit from common mode
rejection wherein changes in vibration frequency of one quartz beam in
response to such
environmental factors as temperature and pressure are negated by equal changes
in
vibration frequency of the other quartz beam.
Quartz beam accelerometers have several disadvantages. For example, in
assembling a quartz beam accelerometer, the quartz beams are typically bonded
or glued
between a support and a seismic mass thereby creating undesirable stresses and
other
problems resulting from thermal expansions. These stresses and problems
adversely
affect the performance of the accelerometer. Moreover, quartz beams normally
have a
high Q when operating in a vacuum. However, when used in an accelerometer,
such
quartz beams often are required to operate in a chamber where a level of gas
pressure is
usually maintained for the purpose of gas damping the seismic mass suspension
CA 02213245 1997-08-18
WO 96/28735 1'CT/US95/03011
3
structure. Unfortunately, this gas pressure also damps the resonating quartz
beams
which thereby decreases the Q, and, therefore, the stability, of the vibrating
quartz
frequency. (The quantity Q as used herein is a quality factor relating to the
stability of a
vibrating device; that is, Q is generally defined as one-half of the kinetic
and potential
energy stored in a vibrating beam divided by the energy lost by a vibrating
beam per
vibration cycle. If the energy applied by a force to the vibrating beam at a
given point
in time is equal to the total energy (i.e. the sum of the kinetic and
potential energies of
the beam) stored in the beam at that point in time, the vibrating beam has no
loss;
however, any difference between this applied energy and the total energy of
the
vibrating beam is the energy lost by the vibrating beam.) Furthermore, in
response to
acceleration, the frequency of the quartz beam can change by only
approximately 10%
of its base frequency through its useful range, i.e. a quartz beam having a
base
frequency of 40,000 Hz., for example, is limited to a 4,000 Hz. variation in
response to
acceleration.
SUMMARY OF THE INVENTION
The resonating silicon beam accelerometer of the present invention allows
silicon processing techniques to be employed in fabricating the silicon beam
accelerometer and thus avoids many of the disadvantages and complexities of
prior art
quartz and silicon beam accelerometers. The resonating silicon beam
accelerometer
may be constructed using cost effective photolithographic techniques common to
the
silicon industry and results in an essentially monolithic structure. Thus,
there are no
glued or bonded joints which create stresses on the vibrating silicon sensing
beams due
to thermal expansion. The construction of the resonating silicon begun
accelerometer of
the present invention permits gas damping of the seismic mass suspension
structure of
the accelerometer while at the same time permitting the silicon beam to
vibrate in a
vacuum chamber such that there is no interference between the use of a damping
gas for
the seismic mass suspension system and the Q of the vibrating silicon beam.
Accord-
ingly, the Q of the vibrating silicon beam can be maintained at a high level.
Furthermore, the vibration frequency of the silicon beam of the resonating
silicon beam
accelerometer of the present invention can vary by over 300% of its base
frequency in
response to acceleration.
CA 02213245 1997-08-18
WO 96/28735 PCT/LTS95/03011
4
An accelerometer according to one aspect of the present invention includes
first,
second, third, and fourth silicon layers wherein each of the silicon layers
has first and
second surfaces. The second surface of the second silicon layer and the first
surface of
the third silicon layer are recessed to form first and second cavities and an
acceleration
responsive mass between the first and second cavities. The first and second
cavities
define (a) first and second silicon flexure members in the first surface of
the second
silicon layer, respectively and (b) third and fourth silicon flexure members
in the second
surface of the third silicon layer respectively. The first silicon layer has
first and second
recesses in the second surface thereof, the first recess being arranged so
that the first
flexure member separates the first cavity and the first recess, and the second
recess
being arranged so that the second flexure member separates the second cavity
and the
second recess. The fourth silicon layer has third and fourth recesses in the
first surface
thereof, the third recess being arranged so that the third flexure member
separates the
first cavity and the third recess, and the fourth recess being arranged so
that the fourth
flexure member separates the second cavity and the fourth recess. A first
means senses
bending of the first flexure member, and a second means senses bending of the
fourth
flexure member.
An accelerometer according to another aspect of the invention includes first,
second, third, and fourth silicon layers wherein each of the silicon layers
has first and
second surfaces. The second surface of the second silicon layer and the first
surface of
the third silicon layer are recessed to form a first continuous cavity
surrounding an
acceleration responsive mass. The first continuous cavity forms a first
flexure member
in the first surface of the second silicon layer and a second flexure member
in the
second surface of the third silicon layer. The acceleration responsive mass is
arranged
to bend the first and second flexure members in response to acceleration. The
first
silicon layer has a second continuous cavity in the second surface thereof.
The second
continuous cavity is arranged so that the first flexure member separates the
first and
second continuous cavities. The fourth silicon layer has a third continuous
cavity in the
first surface thereof. The third continuous cavity is arranged so that the
second flexure
member separates the first and third continuous cavities. First and second
sensing
means sense bending of the first and second flexure members respectively, and
the first
CA 02213245 1997-08-18
WO 96/28735 1'CTlUS95/03011
and second sensing means are arranged with respect to one another for
achieving
common mode rejection.
An accelerometer according to a further aspect of the invention includes a
first
vibrating beam supported by a first flexure member, and a second vibrating
beam
5 supported by a second flexure member. An acceleration responsive :mass is
arranged to
bend the first and second flexure members in at least first and second
directions in
response to accelerations. A damping means damps the first and second flexure
members and the acceleration responsive mass. The damping means includes a
gas.
Vacuum chambers are provided within which the first and second vibrating beams
vibrate. The vacuum chambers are isolated from the gas. First and second
sensing
means sense vibration of the first and second vibrating beams, respectively.
A silicon accelerometer according to a still further aspect of the invention
includes a first silicon vibrating beam supported by a first silicon flexure
member, and a
second silicon vibrating beam supported by a second silicon flexure member. An
acceleration responsive silicon mass is arranged to bend the first and. second
silicon
flexure members in at least first and second directions in response to
accelerations. A
damping means surrounds the first and second silicon flexure members and the
acceleration responsive mass in order to damp the acceleration responsive
silicon mass.
The damping means includes a gas. Vacuum chambers are provided. within which
the
first and second silicon vibrating beams vibrate, the vacuum chambers being
isolated
from the gas. Sensing means sense vibration of the first and second silicon
vibrating
beams, respectively.
BRIEF DESCRIPTION OF TI-IE DRAWING
These and other features and advantages will become more apparent from a
detailed consideration of the invention when taken in conjunction wiuth the
drawing
having the following figures.
Figure 1 shows a prior art push-pull quartz beam accelerometer.
Figure 2 shows another~prior art push-pull quartz beam accelerometer.
Figure 3 shows a prior art double ended quartz beam tuning i;ork
accelerometer.
Figure 4 shows another prior art accelerometer.
Figures 5, 6 and 7 show a known silicon sensor which forms the basis of the
accelerometer according to the present invention.
CA 02213245 1997-08-18
WO 96/28735 PCT/US95/03011
6
Figure 8 is a cross-sectional side view of an accelerometer according to the
present invention.
Figure 9 is a top view of one of the mass/suspension/beam (m/s/b) layers of
the
accelerometer shown in Figure 8.
Figure 10 is a cross-sectional side view taken along lines 10-10 of Figure 9.
Figure 11 shows a circuit arrangement for driving the vibrating silicon beams
of
the accelerometer shown in Figure 8 and for detecting the vibration frequency
thereof.
Figure 12 shows an alternate m/s/b layer for an accelerometer according to the
present invention.
Figure 13 shows still another alternate m/s/b layer for an accelerometer ac-
cording to the present invention.
Figures 14a-c reveal various three wafer resonant microbeam accelerometers.
Figure 15 is a plan view of the middle layer, including the proof mass, of a
resonant microbeam accelerometer.
Figure 16a-d shows several pendulous accelerometers.
Figure 17 is a diagram of electrostatic rebalance electronics for a resonant
microbeam accelerometer.
DETAILED DESCRIPTION
A prior art push-pull quartz beam accelerometer 10 is shown in Figure 1 and
includes an active plate 12 supported between a top support plate 14 and a
lower support
plate 16. Located on the active plate 12 is a seismic mass 18. The upper
support plate
14 has a central opening 20 in order to receive the seismic mass 18. Top
support plate
14 is arranged to limit the travel of the seismic mass 18 along the input
(i.e. sensitive)
axis depicted by a directional arrow 22. The active plate 12 is formed of a
quartz
material having a first opening 24 and a second opening 26 defining a first
quartz beam
28 therebetween. Similar openings on an opposite side of the active plate 12
define a
second quartz beam 30. A mounting bar 32 is included in association with the
lower
support plate 16 in order to mount the quartz beam accelerometer 10 to a
platform.
a
The quartz beams 28 and 30 are driven (by a driver not shown) into vibration
at
a predetermined frequency. As acceleration is experienced by the quartz beam
accelerometer 10, the seismic mass 18 responds to this acceleration by pushing
one of
the quartz beams 28 and 30 into compression and pulling the other of the
quartz beams
CA 02213245 1997-08-18
WO 96/28735 :PCT/US95/03011
7
28 and 30 into tension. The vibration frequencies of the quartz beams 28 and
30 thus
change oppositely to one another and this differential change in vibration
frequency is
sensed in order to provide an indication of the amount of acceleration
experienced by
the push-pull quartz beam accelerometer 10.
Shown in Figure 2 is another prior art dual vibrating quartz beam
accelerometer
34. The dual vibrating quartz beam accelerometer 34 includes a housing 36
which
supports a vibrating beam support assembly 38. A first pendulum 40 is
suspended by
the support assembly 38 by way of a hinge 42 and a second pendulum 44 is
suspended
by the beam support assembly 38 by way of a hinge 46. A first vibrating quartz
beam
48 is suitably attached between the pendulum 40 and the beam support assembly
38 and
a second vibrating quartz beam 50 is suitably attached between the second
pendulum 44
and the beam support assembly 38. The beam support assembly 38 is configured
to
provide gaps 52 and 54 between the first pendulum 40 and the beam support 38
in order
to limit travel of the pendulum 40 in either direction of the input axis of
the
accelerometer 34 depicted by a directional arrow 56. By the same token, the
support
assembly 38 is configured to provide gaps 58 and 60 between the second
pendulum 44
and the beam support 38 in order to limit the travel of the pendulunn 44 in
either
direction along the input axis depicted by the directional arrow 56. The beam
support
assembly 38 also provides gas film damping surfaces that cooperate with the
surfaces of
the pendulums to damp the pendulums 40 and 44.
The quartz beams 48 and 50 are driven into vibration at a base frequency. As
the
accelerometer 34 experiences acceleration along its input axis depicted by the
directional arrow 56, one of the quartz beams 48 and 50 is pushed into
compression by
its corresponding pendulum, and the other of the quartz beams 48 quid 50 is
pulled into
tension by its corresponding pendulum. For example, if the accelerometer 34 is
accelerated in the upward direction as viewed in Figure 2, the pendulums 40
and 44 tend
to rotate downward with respect to the beam support 38. This rotation of the
pendulums
40 and 44 pushes the quartz beam 48 into compression and pulls the quartz beam
50 into
tension. The resulting change in the forces acting on the quartz be~uns 48 and
50
differentially alters the vibration frequencies of the quartz beams 48 and 50.
That is, as
in the case of the accelerometer 10 shown in Figure 1, the vibration frequency
of the
quartz beam 48 decreases and the vibration frequency of the quartz; beam 50
increases.
CA 02213245 1997-08-18
WO 96/28735 PC'T/US95/03011
8
This differential frequency change is sensed by oscillators 64 and 66 in order
to provide
an indication of the amount of acceleration experienced by the accelerometer
34.
Figure 3 illustrates a prior art double-ended tuning fork quartz beam
accelerometer 70. The accelerometer 70 includes a housing 72 containing a
seismic
mass 74 attached to the housing 72 by a hinge 76. A first quartz beam 78 spans
the
hinge area 76, on one side of the seismic mass 74, from the housing 72 to the
seismic
mass 74. A second quartz beam 80 similarly spans the hinge area 76, on the
opposite
side of the seismic mass 74, from the housing 72 to the seismic mass 74. The
housing
72 and the seismic mass 74 are configured to provide a pair of gaps 82 and 84
between
the seismic mass 74 and the housing 72. These gaps act as shock caging gaps in
order
to limit travel of the seismic mass 74 resulting from accelerations along the
input axis of
the accelerometer 70 depicted by a directional arrow 86. The housing 72 also
provides
gas film damping surfaces that cooperate with the surfaces of the seismic mass
to damp
the seismic 74.
The quartz beams 78 and 80 are typically driven at a base vibration frequency
by
a source of electrical energy. As the seismic mass 74 pivots about the hinge
72 in
response to an acceleration along the input axis depicted by the directional
arrow 82, the
seismic mass 74 pivots about the hinge 76 which causes the seismic mass 74 to
assume
the position shown by the dotted line 88 in Figure 3. If the movement of the
seismic
mass 74 is downward as shown in Figure 3, the quartz beam 78 is stretched by
an
amount 90, and the quartz beam 80 is compressed by a similar amount. As the
seismic
mass 74 so pivots in either direction along its input axis, the seismic mass
74 pushes one
of the quartz beams 78 and 80 into compression and pulls the other of the
quartz beams
78 and 80 into tension. This change in the tension and compression of the
quartz beams
78 and 80 results in a differential change of the vibration frequencies of the
quartz
beams 78 and 80. This differential change of the vibration frequencies of the
quartz
beams 78 and 80 provides an indication of the amount of the acceleration.
As shown in Figure 4, another prior art accelerometer 100 includes a housing
102 and a cover 104. Within the housing 102 is a seismic mass 106 suspended by
seismic mass supports 108 and 110. A first quartz beam 112 is attached at one
end to
the seismic mass support 108 and at the other end to the seismic mass 106. A
second
quartz beam 114 is attached at one end to the seismic mass support 110 and at
the other
CA 02213245 1997-08-18
WO 96/28735 I'CTlUS95103011
9
end to the seismic mass 106. As the accelerometer 100 experiences
accelerations along
its input axis depicted by a directional arrow 116, the seismic mass :106
moves to push
one of the quartz beams 112 and 114 into compression and pull the other of the
quartz
beams 112 and 114 into tension.
The quartz beams 112 and 114 may be electrically vibrated at a base frequency.
As the accelerometer 100 experiences acceleration, the seismic mass 106 moves
to push
one of the quartz beams 112 and 114 into compression and to pull the other of
the quartz
beams 112 and 114 into tension which results in a differential change in the
vibration
frequencies of the quartz beams 112 and 114. This differential vibration
frequency
change provides an indication of the amount of acceleration experienced by the
accelerometer 100.
Such prior art accelerometers employ either stiffened flexure; members to
support
the seismic mass or a damping gas in order to avoid undesirable sensitivities
to
vibration. If stiffened flexure members are employed, the sensitivir,~ of the
accelerometer is traded off against the desired lower frequency response of
the
accelerometer. If a damping gas is used, the internal volume of the
.accelerometer
housing is filled with a damping gas at a pressure which is selected to
provide the proper
amount of damping for the seismic mass or the pendulum of the accelerometer.
As
discussed above, this gas has the unfortunate result that, not only is 'the
seismic mass or
pendulum damped, but the vibrating quartz beams are also damped. Damping of
the
vibrating quartz beams decreases the Q of the accelerometer. A decrease in Q
reduces
the frequency stability of the vibrating quartz beams.
Furthermore, the accelerometers of Figures 1 - 4 are mechanically complex
assemblies which are costly to manufacture and to assemble. The quartz beams
included in these assemblies are typically bonded or glued into place which
creates
undesirable stresses and thermal expansion problems. Furthermore, a quartz
beam,
s
because it typically has a low tensile strength, has a limited range within
which its
vibration frequency can change. For example, a quartz beam havint; a base
vibration
frequency of 40,000 hertz is limited to a range of 4,000 hertz centered about
its base
frequency. A vibrating quartz beam will break if it is vibrated at a frequency
above the
upper frequency of this frequency range.
CA 02213245 1997-08-18
WO 96/28735 PCT/US95/03011
On the other hand, a silicon beam has a higher tensile strength and,
therefore,
can undergo a greater range of vibration frequency variation. Thus, a silicon
beam can
experience frequency variations on the order of 300% of the base frequency.
When a
r
vibrating silicon beam is used in an accelerometer, this greater range
produces better
5 resolution of the output signal which provides the indication of
acceleration. A fiirther
advantage of using silicon vibrating beams is that the silicon accelerometer
can be
fabricated using cost effective photolithographic techniques common to the
silicon
industry.
A silicon sensor 130, which may be used in the accelerometer of the present
10 invention, is shown in Figures 5, 6 and 7. The silicon sensor 130 may be
made by one
or more of the processes shown in the following U.S. Patents: 4,744,863;
4,853,669;
4,897,360; 4,996,082; 5,013,693; and, 5,090,254. The silicon sensor 130
includes a thin
silicon beam 132 of polysilicon material which may be deposited on a
semiconducting
crystalline silicon flexure member 134. The thickness of the silicon beam 132
may be
on the order of two micrometers, for example. A vacuum chamber 136 is formed
partially beneath the silicon beam 132. The vacuum chamber 136, together with
a pair
of slots 138 and 140, permits the silicon beam 132 to vibrate with a very high
Q.
A drive capacitor 141 can be used to vibrate the silicon beam 132 at a base
frequency. The drive capacitor 141 can be either a two electrode capacitor or,
preferably, a three electrode capacitor. Only one capacitor electrode 142 of
the drive
capacitor 141 is shown in Figures 5 and 7. The electrode 142 may be a metal
film
suitably adhered to one of the surfaces of the silicon beam 132 or,
preferably, the
electrode 142 may be a doped region on the surface of the silicon beam 132. A
second
electrode (not shown) of the drive capacitor 141 can be similarly doped into
the flexure
member 134. A cover 146 is placed over at least a portion of the silicon beam
132 to
complete the vacuum chamber 136 within which the silicon beam 132 vibrates.
The
cover 146 should have an interior recess of sufficient depth to permit the
silicon beam
132 to vibrate in response to the drive capacitor 141. A third electrode (not
shown) can
be supported by the cover 146. For example, if the cover 146 is a silicon
cover, the
third electrode may be doped into a surface of the cover 146 which faces the
silicon
beam 132. Finally, the silicon sensor 130 also includes a sensing
piezoelectric resistor
148 which senses the vibration of the silicon beam 132 and which may be doped
into the
CA 02213245 1997-08-18
WO 96/28735 IPCT/US95/03011
11
same surface of the silicon beam 132 into which the electrode 142 is doped.
The
piezoelectric resistor 148 has terminals 150 and 152.
In using the silicon sensor 130 in the accelerometer of the present invention,
an
alternating current may be supplied to the drive capacitor 141 in order to
vibrate the
silicon beam 132. If the drive capacitor 141 is a two electrode capa<;itor, an
alternating
current energizes the two electrodes of the drive capacitor 141 to produce an
alternating
electrostatic field in order to vibrate the silicon beam I32 at the frequency
of the
alternating current. If the drive capacitor 141 is a three electrode
capacitor, an
alternating current is applied across the upper capacitor electrode on the
cover 146 and
the lower capacitor electrode on the flexure member 134 so that an alternating
electrostatic field exerts an alternating force on the silicon beam 132 to
vibrate the
silicon beam 132 at the frequency of the alternating current. The middle
electrode 142
is connected to circuit ground.
The inputs of an oscillator amplifier (not shown in Figures 5--7) can be
connected
to the piezoelectric resistor 148 by way of the terminals 150 and 152, and the
outputs of
this oscillator amplifier can be connected to the electrodes of the drive
capacitor 141
which vibrates the silicon beam 132. The resonant frequency at which the
silicon beam
132 vibrates is determined by the strain on the silicon beam 132. Accordingly,
the
resistance value of the piezoelectric resistor I48 changes with changes in the
resonant
frequency of the silicon beam 132. Thus, the piezoelectric resistor 148
provides a signal
to the oscillator amplifier wherein the frequency of this signal is the
resonant frequency
of the silicon beam 132. The amplifier provides an output to the capacitor 141
wherein
the frequency of the output from the amplifier is the resonant frequency of
the silicon
beam 132. The drive capacitor 141 provides sufficient energy to sustain the
vibration of
the silicon beam 132.
Thus, the silicon sensor 130 can be used to sense forces which result in
bending
of the flexure member 134. That is, as a force causes the flexure mennber 134
to bend as
shown in Figure 6 so that the silicon beam 132 of the silicon sensor 1L 30
lies on the
outside surface of the curvature of the flexure member 134, the silicon beam
132
experiences an increase in tension causing its resonant frequency to increase.
Likewise,
as a force causes the flexure member 134 to bend so that the silicon beam 132
lies on
CA 02213245 1997-08-18
WO 96/28735 PCT/US95/03011
12
the inside surface of the curvature of the flexure member 134, the silicon
beam 132
experiences an increase in compression causing its resonant frequency to
decrease.
The change in output frequency of the oscillator amplifier can be sensed as an
Y
indication of the amount, and the direction, of forces which are applied to
the flexure
member 134. Typical silicon beams formed on silicon flexures may have an
unstressed
base resonant frequency of from 500,000 hertz to 600,000 hertz. Under tension,
this
frequency swings as high as 2,000,000 hertz, and under compression this
frequency
swings as low as 150,000 hertz. This frequency range is a very large span of
useful
output, and can provide high resolution and high accuracy in many force
sensing
applications.
A pair of silicon sensors 130 can advantageously be used in a silicon
accelerometer 160 shown in Figure 8. The silicon accelerometer 160 is
constructed of
four crystalline silicon plates or layers 162, 164, 166 and 168. The outer
layers 162 and
168 act as cover layers and may be of identical geometric construction with
respect to
one another. The middle layers 164 and 166 are mass/suspension/beam layers and
also
may be of identical geometric construction with respect to one another.
Anisotropic
etching may be employed to form the recesses, the suspension flexures, and the
seismic
mass in the cover layers 162 and 168 and in the m/s/b layers 164 and 168, as
shown,
before the layers 162, 164, 166, and 168 are assembled to form the
accelerometer 160.
As viewed in Figure 8, each of the layers 162-168 has an upper surface, i.e. a
first surface, and a lower surface, i.e. a second surface. The first surface
of the m/s/b
layer 166 and the second surface of the m/s/b layer 164 may be etched or
otherwise
suitably processed to provide cooperating recesses 170 and 172 which are
arranged to
form a cavity 174 when the m/s/b layers 164 and 166 are assembled as shown in
Figure
8. Similarly, the first surface of the m/s/b layer 166 and the second surface
of the m/s/b
layer 164 may be etched or otherwise suitably processed to provide cooperating
recesses
176 and 178 which are arranged to form a cavity 180. The cover layer 162 is
provided
with recesses 182 and 184 and a travel limit stop 186, and the cover layer 168
is
provided with recesses 188 and 190 and a travel limit stop 191.
Sufficient silicon material remains in the second surface of the m/s/b layer
166
after formation of the recesses 170 and 176 to form flexure members 192 and
194,
respectively. Similarly, sufficient silicon material remains in the first
surface of the
CA 02213245 1997-08-18
WO 96/28735 1'CTIUS95/03011
13
m/s/b layer 164 after formation of the recesses 172 and 178 to form flexure
members
196 and 198, respectively. A first silicon sensor 130, including a silicon
beam 132, a
drive capacitor 141, a piezoelectric resistor 148, and a cover 146, is
provided in a region
200 of the flexure member 192 of the same construction as the silicon sensor
130 shown
in Figures 5-7. Thus, the flexure member 192 is the same as the flexure member
134 of
the silicon sensor 130 of Figures 5-7. Similarly, a second silicon sensor 130,
including
a silicon beam 132, a drive capacitor 141, a piezoelectric resistor 14.8, and
a cover 146,
is provided in a region 202 of the flexure member 198 of the same construction
as the
silicon sensor 130 shown in Figures S-7. Thus, the flexure member 198 is also
the same
as the flexure member 134 of the silicon sensor 130 of Figures 5-7.
A terminal pad 204 is provided on the second surface of the m/s/b layer 166
and
suitable circuit paths on the second surface of the m/s/b layer 166 connect
the terminal
pad 204 to the drive capacitor 141 and the piezoelectric resistor 148 provided
in the
region 200 of the m/s/b layer 166. A similar terminal pad 206 is provided on
the first
surface of the m/s/b layer 164 and suitable circuit paths on the first surface
of the m/s/b
layer 164 connect the terminal pad 206 to the drive capacitor 141 and
piezoelectric
resistor 148 provided in the region 202 of the m/s/b layer 164. As v~ill be
discussed
below, the terminal pads 204 and 206 provide the means for connecting the
piezoelectric
resistors 148 and the drive capacitors 141 in the regions 200 and 20:? to
respective
oscillator amplifiers.
When the m/s/b layers are assembled as shown in Figure 8, i.he recesses 170
and
176 in the m/s/b layer 166 and the recesses 172 and 178 in the m/s/lr layer
164 form a
seismic mass 208 between the cavities 174 and 180. The accelerometer 160 is
assembled by bonding the layers 162, 164, 166 and 168 together using silicon
0
dioxide/silicon thermoelectric bonds. As the accelerometer 160 experiences
acceleration, the seismic mass 208 bends the flexure members 192 a:nd 198 so
that one
of the silicon beams 132 in the regions 200 and 202 is on the outside surface
of its
corresponding flexure member and so that the other of the silicon beams 132 in
the
regions 200 and 202 is on the inside surface of its corresponding flexure
member. Thus,
one silicon beam 132 is in tension and the other silicon beam 132 is in
compression.
For example, if the accelerometer 160 is accelerated in the upward dLirection
as viewed
in Figure 8, the seismic mass 208 bends the flexure member 192 so that the
silicon beam
CA 02213245 1997-08-18
WO 96/28735 PCT/US95/03011
14
132 in the region 200 is on the inside surface of the curvature of the flexure
member 192
and is, accordingly, in compression, and the seismic mass 208 bends the
flexure member
198 so that the silicon beam 132 in the region 202 is on the outside surface
of the
curvature of the flexure member 198 and is, accordingly, in tension. Thus, the
resonant
frequencies of the silicon beams 132 in the regions 200 and 202 are decreased
and
increased, respectively. This differential change in the vibration frequencies
may be
sensed to provide an indication of the amount and direction of the
acceleration.
The cavities 174 and 180, together with the recesses 182, 184, 188, and 190,
may be filled with a damping gas in order to damp movement of the seismic mass
208
and to thereby decrease the sensitivity of the accelerometer 160 to vibration.
Although
many gases may be used for the damping gas in the cavities 174 and 180 and the
recesses 182, 184, 188, and 190, it is preferable that the gas be inert and
thermally
conductive. The vacuum chambers 136 of the silicon sensors 130 in the regions
200 and
202, i.e. the vacuum chambers within which the silicon beams 132 vibrate,
ensure a high
Q for the accelerometer 160. Preferably, the cavities 174 and 180 and the
recesses 182,
184, 188 and 190 are in gas communication with one another through suitable
holes (not
shown in Figure 8) in the flexure members 192, 194, 196, and 198. The seismic
mass
208 may move between the cover layers 162 and 168 but the extent of movement
of the
seismic mass 208 is limited by the travel limit stops 186 and 191,
respectively.
Although the accelerometer 160 may be an elongated device having separate
cavities 174 and 180, the accelerometer 160 is preferably a quadrilateral
device.
Accordingly, although only the top view of one of the layers 162, 164, 166,
and 168 is
shown in Figure 9 (i.e. the m/s/b layer 164), it is understood that the other
layers 162,
166 and 168 have the same quadrilateral shape. The first m/s/b layer 164 as
shown in
Figures 9 has two recesses 220 and 222 joining the recesses 174 and 178 to
form one
continuous recess 224. The continuous recess 224 is etched as a closed path in
the
second surface 226 of the first m/s/b layer 164. This etching results in a
mesa 228
0
surrounded by the continuous recess 224. A similar mesa 230 (Figure 8) is
formed in
the m/s/b layer 166. When the m/s/b layers 164 and 166 are assembled together
as
shown in Figure 8 (such as by bonding a ridge 232 around an outside perimeter
of the
layer 164 to a ridge 234 around an outside perimeter of the layer 166), the
mesa 228 in
CA 02213245 1997-08-18
W O 96128735 FCT/LTS95/03011
1$
the m/s/b layer 164 and the corresponding mesa 230 in the m/s/b layer 166 form
the
seismic mass 208 shown in Figure 8.
Holes 236 may be provided from the continuous recess 224 through a first
surface 238 of the m/s/b layer 164 in order to provide for communication
between the
continuous recess 224 and the recesses 182 and 184 in the cover layer 162.
Holes
similar to the holes 236 may also be provided from the similar continuous
recess in the
m/s/b layer 166 through a first surface thereof in order to provide for
communication
between that continuous recess and the recesses 188 and 190 in the cover layer
168.
Since the cover layers 162 and 168 are also quadrilateral, the recesses 182
and 184 form
a continuous quadrilateral recess in the second surface of the cover layer
162, and the
recesses 188 and 190 form a continuous quadrilateral recess in the first
surface of the
cover layer 168. Thus, the cavities 174 and 180, the recesses 182, 184, 188,
and 190,
and the holes 236 form one continuous gas chamber for containing a damping gas
to
damp movement of the seismic mass 208 and the flexure members 192 and 198.
The terminal pad 206 may be provided on the m/s/b layer 164 having terminals
240, 242, 244 and 246 for connection to the drive capacitor 141 and
:piezoelectric
resistor 148 of the region 202 on the surface 238 of the m/s/b layer 164. The
region 250
shown in Figure 9 represents the drive capacitor 141/piezoelectric resistor
148 region of
the silicon beam 132 in the region 202 on the m/s/b layer 164. The terminal
240 of the
terminal pad 206 is connected to one electrode of a two electrode drive
capacitor 141,
and the terminal 246 is connected to the other electrode of the drive
capacitor 141 (it
being understood that an additional terminal in the terminal pad 206 is
required if the
drive capacitor 141 includes a third electrode). The terminal 242 is connected
to the
terminal 150 of the piezoelectric resistor 148 in the region 202 on the; m/s/b
layer 164,
and the terminal 244 is connected to the terminal 152 of this piezoelectric
resistor 148.
Similar terminals of the terminal pad 204 on the m/s/b layer 166 are connected
to the
drive capacitor 141 and the piezoelectric resistor 148 in the region 2010 on
the m/s/b
layer 166.
An oscillator amplifier circuit 290 is connected to the accelerometer 160 as
shown in Figure 11 in order to drive the silicon beams 132 in the regions 200
and 202
into vibration and to sense the resonant frequency of these silicon beams 132.
This
oscillator amplifier circuit 290 includes an oscillator amplifier 300 having
its inputs 302
CA 02213245 1997-08-18
R'O 96/28735 PCTIUS95/03011
16
and 304 connected to the piezoelectric resistor 148 of the region 202 by way
of the
terminals 240 and 246. A first output 306 from the oscillator amplifier 300 is
connected
to one electrode of the drive capacitor 141 formed in the region 202 by way of
the
terminal 244. A second output 308 from the oscillator amplifier 300 is
connected to
ground and also to the other electrode of the drive capacitor 141 in the
region 202 by
way of the terminal 242. The output 306 from the amplifier 300 is also
connected to a
first input of an amplifier 310 and may be connected to a first input of an
amplifier 312.
Similarly, the oscillator amplifier circuit 290 also includes an oscillator
amplifier
318 having its inputs 314 and 316 connected to the piezoelectric resistor 148
of the
region 200 by way of the terminals 242' and 244'. A first output 320 from the
oscillator
amplifier 318 is connected to one electrode of the drive capacitor 141 formed
in the
region 200 by way of the terminal 240'. A second output 322 from the
oscillator
amplifier 318 is connected to ground and also to the other electrode of the
drive
capacitor 141 in the region 200 by way of the terminal 246'. The output 322
from the
amplifier 318 is also connected to a second input of the amplifier 310 and may
be
connected to a second input of the amplifier 312.
During resonance of the silicon beam 132 in the region 202 of the
accelerometer
160, the oscillator amplifier 300 supplies alternating current to the drive
capacitor 141 in
the region 202 to vibrate the silicon beam 132 in this region at its resonant
frequency.
This frequency is sensed by the piezoelectric resistor 148 in the region 202
which
provides an input to the oscillator amplifier 300 in order to maintain
resonant vibration
of the silicon beam 132 in the region 202. Similarly, during resonance of the
silicon
beam 132 in the region 200 of the accelerometer 160, the oscillator amplifier
318
supplies alternating current to the drive capacitor 141 in the region 200 to
vibrate the
silicon beam 132 in this region at its resonant frequency. This frequency is
sensed by
the piezoelectric resistor 148 in the region 200 which provides an input to
the oscillator
amplifier 318 in order to maintain resonant vibration of the silicon beam 132
in the
region 200.
As the accelerometer 160 experiences acceleration, the seismic mass 208 bends
the flexure members 192 and 198 to place one of the silicon beams 132 in the
regions
200 and 202 in tension and the other of the silicon beams 132 in the regions
200 and
202 in compression. The resonant frequency of one of the silicon beams 132 in
the
CA 02213245 1997-08-18
WO 96/28735 IPCT/L1S95/03011
17
regions 200 and 202 increases, and the resonant frequency of the other of the
silicon
beams 132 in the regions 200 and 202 decreases. Accordingly, the output
frequency
from one of the oscillator amplifiers 300 and 318 increases and the output
frequency
from the other of the oscillator amplifiers 300 and 318 decreases. The outputs
306 and
308, and 320 and 322, respectively, from the oscillator amplifiers 300 and
318, are fed
back to their respective drive capacitors 141 in the regions 200 and :Z02 to
sustain the
new resonant frequencies of the silicon beams 132. This process maintains the
resonant
vibration frequencies of the silicon beams 132 in the regions 200 and 202
during
changes in acceleration experienced by the accelerometer 160.
Thus, the output frequencies from the oscillator amplifiers 300 and 318 change
in opposite directions with one increasing and the other decreasing. The
amplifier 310
provides an output 324 which is a function of the difference between the
output
frequencies from the oscillator amplifiers 300 and 318. Thus, the output 324
from the
amplifier 310 indicates the amount of acceleration which produced i:his
difference
between the output frequencies from the oscillator amplifiers 300 arid 318.
The polarity
of the frequency change in the output 324 of the amplifier 310 indicates the
direction of
acceleration. The amplifier 312 provides at output 326 a frequency which is a
function
of the sum of the output frequencies from the oscillator amplifiers 300 and
318. The
output 326 from the amplifier 312 can provide an indication of temperature and
other
factors which influence the vibration frequencies of the silicon beams 132 in
the regions
200 and 202 and, if desired, can be used for compensation.
The construction of the accelerometer 160 uses cost effective:
photolithographic
techniques common to the silicon industry and results in an essentially
monolithic
structure. The performance of the accelerometer 160 is enhanced because the
resonant
beam structure, i.e. the vibrating beams 132, are sealed in their own vacuum
compartments. Thus, any interference between the resonant beam ~1 and the use
of a
damping gas for the seismic mass 208 and the flexure members 192 and 198 is
avoided.
The accelerometer 160 is very sensitive along its input axis 210 while
providing a stiff
suspension along the directions perpendicular to the input axis 210. The
arrangement of
the accelerometer 160 provides for common mode rejection in that environmental
influences negate one another by the use of dual vibrating silicon beams.
Also, the
arrangement shown in Figure 8, and in particular the geometry shov~m in Figure
9, can
CA 02213245 1997-08-18
WO 96/28735 PCTIL1S95/03011
18
accommodate multiple pairs of silicon beams 132. Thus, a silicon beam 130 can
be
placed on each side of the four sided structure shown in Figure 9. Such an
arrangement
can provide for redundancy in the event of failure. The temperature range of
the device
is limited only by the active electronics and is not limited by accelerometer
itself since
the use of polysilicon avoids the need to form pn junctions in the doped
areas.
An alternate m/s/b layer 400 is shown in Figure 12. This m/s/b layer 400 may
be
used in place of the m/s/b layers 164 and 166 shown in Figures 8 and 9. The
m/s/b layer
400 has a quadrilateral continuous recess 402. The continuous recess 402 is
etched as a
closed path in a surface 404 of the m/s/b layer 400. This etching results in a
mesa 406
surrounded by the continuous recess 402. As in the case of the m/s/b layers
164 and 166
of Figures 8 and 9, the mesa 406 in the m/s/b layer 400 forms a seismic mass
with a
corresponding mesa in a corresponding second m/s/b layer.
As shown in Figure 12, the continuous recess 402 is selectively etched
entirely
through the m/s/b layer 400 in multiple areas 408 to leave webs 410 physically
attaching
and supporting the mesa 406 to a ridge 412 around the perimeter of the m/s/b
layer 400.
As in the case of the holes 236 in the m/s/b/ layer 164, the areas 408 provide
for gas
communication through the continuous recess 402. Opposing ones of the webs 410
act
as flexure members similar to the flexure member 134 of Figure 5. Silicon
sensors may
thus be formed in these opposing webs. The size of webs 410 can be controlled
to in
turn control the amount of stiffness with which the mesa 406 is supported and
the
stiffness of the flexure member formed by the webs 410 of the m/s/b layer 400.
Accordingly, the sensitivity of the accelerometer can be increased while at
the same
time the mass of the accelerometer is reduced.
A further alternate m/s/b layer 500 is shown in Figure 13. The m/s/b layer 500
has a circular continuous recess 502. Whereas a quadrilateral recess may be
formed by
anisotropic etching, the circular recess 502 is formed by isotropic etching.
This etching
of the recess 502 results in a circular mesa 504 surrounded by the continuous
recess 502.
As in the case of the m/s/b layers 164 and 166 of Figures 8 and 9, the mesa
504 in the
m/s/b layer 500 forms a seismic mass with a corresponding mesa in a
corresponding
second m/s/b layer. This second m/s/b layer may be formed having the same
geometry
as the m/s/b layer 500. Two additional layers, similar to the outer layers 162
and 168
shown in Figure 8, may also be formed, but with circular geometry to match the
circular
CA 02213245 1997-08-18
WO 96/28735 l'C'TIUS95/03011
19
geometry of the m/s/b layer 500. As in the case of the holes 236 in the m/s/b/
layer 164,
gas holes 506 and 508 may be provided.
Figure 14a shows a configuration 600 having only three silicon wafers or
layers
601, 602 and 603. Figure 15 is a top view of second layer 602. Wafer 602 is
recessed
at volumes 609 and 610 to form a proof mass or an acceleration res~~onsive
mass 604
supported by flexure members 605, 606, 607 and 608 to supportive structure
611. Other
supporting flexure members 616 are show in figure 15. When subject to
acceleration in
directions 612 and 613, mass 604 moves towards layer 601 or 603, respectively.
Recesses 614 and 61 S permit movement of mass 604.
Recesses 614 and 615 have damping channels 617 and dimples 618 for damping.
Layers 601 and 603 can be used to seal the chamber that the proof mass is in
for
purposes of containing a damping gas or maintaining a vacuum. Layers 601 and
603
also provide for overrange and environmental protection.
Silicon resonant beam sensors 620 and 622 incorporate silicon or polysilicon
beams 131 and 132, respectively, and associated sensing and drive elements as
noted
above. Pads 619 provide for external connections to sensors 620 ar.~d 622.
Sensors 620
and 622 in figure 14a are placed on flexure members 605 and 606, respectively.
Location of sensors 620 and 622 on flexure members is significant. Sensor 620
is
placed on flexure member 605 closer to supportive structure 611 th;~n to proof
mass 604.
On the other hand, sensor 622 is placed on flexure member 606 closer to proof
mass 604
than to supportive structure 616. When proof mass moves in direction 612, beam
131 of
sensor 620 is in compression and beam 132 of sensor 622 is in tension thereby
resulting
in resonant beam frequencies that are indicative of the respective compression
and
tension. These frequencies reveal the amount of movement of mass 604 towards
wafer
601 and consequently the amount and direction of acceleration force exerted on
device
600. When proof mass moves in direction 613, beam 131 of sensor 620 is in
tension
and beam 132 of sensor 622 is in compression thereby resulting in resonant
beam
frequencies that are indicative of the respective tension and compression. A
resonant
beam frequency increases with tension and decreases with compression. The
frequencies reveal the amount of movement of 604 into recess 614 or 615 and
consequently the amount and direction of acceleration force exerted on device
600.
CA 02213245 1997-08-18
WO 96/28735 PCT/US95/03011
Figure 14b shows configuration 630 regarding placement of sensor 622. Sensor
620 remains in the same location on flexure member 605, as in configuration
600.
Sensor 622 is also located on flexure member 605 but closer to proof mass 604
than to
supportive structure or frame 611. Sensor 622 is in tension while sensor 620
is in
5 compression and vice versa, in the same manner as device or configuration
600 of figure ,
14a, for respective proof mass 604 displacements, when device 630 is subject
to an
acceleration force in direction 612 or 613, respectively. Devices 600 and 630
are
insensitive to and do not measure acceleration forces in directions
perpendicular to
directions 612 and 613.
10 Figure 14c shows configuration of device 640. Device 640 is similar to
devices
600 and 630 except for a different placement of sensor 622 which is situated
on flexure
607 at a position closer to supportive structure or frame 611 than to proof
mass 604.
Sensor 620 is in tension while sensor 622 is in compression and vice versa, in
the same
manner as device or configuration 600 of figure 14a, for respective proof mass
604
15 displacements when device 640 is subject to an acceleration force in
direction 612 or
613, respectively.
Figure 15 is a top view of device 600 of figure 14a. Top views of devices 630
and 640 are the same except for the different locations of sensor 622 as noted
in figures
14b and 14c. Proof mass 604 is supported by flexure members 616 in addition to
20 flexure members 605 and 606, which are connected to structure 611, at the
top surface
of layer 602. Additional flexure members, besides members 607 and 608, support
mass
604 at the bottom surface of layer 602. Pads 619 provide for external
connections to
sensors 620 and 622, and to layers 601, 602 and 603, as needed, such as in the
electrostatic force rebalance design of figure 17. Figure 15 also shows a
microbeam
sensor 629 which provides a signal indicative of the temperature of layer 602
for
compensating sensors 620 and 622 to reduce or eliminate temperature-related
sensing
deviations.
Figures 16a, 16b, 16c arid 16d reveal configurations or devices 631, 632 and
633
of a pendulous microbeam sensor accelerometer design. Acceleration responsive
mass
or proof mass 634 swings on flexure members 605 and 607, or just on flexure
member
605 which function like supportive hinges for mass 634. Mass 634 stays
equidistant
CA 02213245 1997-08-18
i~VO 96/28735 IPCT/US95/03011
21
from wafers 601 and 603 unless subjected to an acceleration force in direction
612 or
613.
In device 63 I of figure 16a, sensor 620 is placed on flexure Ei05 at a
location
closer to supportive structure 611 than to proof mass 634, and sensor 622 is
placed on
flexure 605 at a location closer to proof mass 634 than to supportive
structure 611.
When an acceleration force acts on proof mass 634 in direction 612, sensor 620
has
microbeam 131, which like microbeam 132, in compression and sensor 622 has
microbeam 132 in compression. When an acceleration force acts on proof mass
634 in
direction 613, sensors 620 and 622 have beams 131 and 132, respectively, in
tension.
Compression and tension affect the resonant frequencies of beams 1:3 l and
132. The
resonant frequencies are indicative of the magnitude and direction oi°
the affecting
acceleration force on proof mass 634.
Configuration or device 632 has sensor 620 placed on flexure member 605 and
situated closer to supportive structure 611 than to proof mass 634, and sensor
622 on
flexure member 607 and situated closer to supportive member 611 than to proof
mass
622. When device 632 is subject to an acceleration force in direction 612 or
613, sensor
620 is in compression while sensor 622 is in tension, and vice versa,
respectively, with
determination of magnitude and direction of the acceleration force iriade from
the
resulting changes of resonant beam frequencies.
Configuration or device 633 has sensor 620 placed on flexure: member 605 and
sensor 622 on supportive structure or frame 611. Sensor 622 is not effected by
an
acceleration force but is by temperature and~may be used for temperature
compensation
of sensor 620. When device 633 is subject to an acceleration force in
direction 612 or
613, resonant beam 131 is under compression or tension, respectively, and has
a
resonant frequency indicative of the compression or tension and, in turn, of
the
magnitude and direction of the acceleration force.
Figure 17 is a diagram of electronics 650 for a electrostatic force rebalance
resonant beam sensor system. Resonant beam sensors 620 and 622 ~~re on flexure
member 605 of device 600. With no accelerational force, the resonant
frequencies from
sensors 620 and 622 are about the same, and have a particular value :for no
force and the
frequencies have a particular phase relationship. Accelerational force in
direction 612
causes beam 131 of sensor 620 to be in compression and beam 132 o~f sensor 622
to be
CA 02213245 1997-08-18
WO 96/28735 PCT/US95/03011
22
in tension. Thus, the resonant frequency of sensor 620 decreases as the
resonant
frequency of sensor 622 increases. Signals 641 and 642, representing resonant
frequencies of sensors 620 and 622, respectively, go to phase discriminator
643 which
detects the changes in the relative phases of the resonant frequencies and
indicates
whether the proof mass 604 is moving towards recess 614 of layer 601, or
moving
towards recess 615 of layer 603. An output indicating a change of phase
relationship
between the resonant frequencies of sensors 620 and 622, goes from phase
discriminator
643 to amplifier 644. The outputs of amplifier 644 go to layers 601 and 602,
respectively. Layer 602 is connected to a reference voltage. Pairs of layers
601 and
602, and 602 and 603, form electrodes for two capacitors between the layers.
Layers
645 and 646 insulate layers 601 and 603 from layer 602. The outputs of
amplifier 644
to device 600 cause an electrostatic force to be present in the device to
force back or
rebalance proof mass 604 to its original mid-position of equilibrium as if no
accelerational force were acting on mass 604. The amount of voltages required
at the
outputs of amplifier 644 to rebalance proof mass 604 indicates the amount and
direction
of accelerational force, perpendicular (i.e., direction 612 or 613) to the
plane of the
surfaces of layers 601, 602 and 603. One output of amplifier 644 goes to the
inverting
input of amplifier 647. Resistor 648 provides negative feedback to and
determines the
gain of amplifier 647 to provide a particular scale factor at the output of
amplifier 647,
which results in a signal capable of providing an accurate indication of the
acceleration
force acing on proof mass 604. A voltage is provided at the non-inverting
input of
amplifier 647 for bias adjustment.
Since variations can be made with respect to the invention without departing
from the scope of the invention, the scope of the invention is to be limited
only by the
claims.
