2.2.

s Pres sure Sens ors

As we learned in Chapter 1 cropressure sensors are widely used in automotive and
aerospace industries. Most these sensors nc 10n on the principle of mec an1ca e-
fonnation and stresses of thin diaphragms induced by the measurand pressure. M~-
chanically induced diaphragm deformation and stresses are then converted into
electrical signal output through several means of transduction.
There are generally two types of pressure sensor: absolute and gage pressure
sensors. The absolute pressure sensor has an evacullied l,av1ty on one s1ae oi the di-
aphragm. The measured pressure is the "absolute' .."value with vacuum as the refer-
ence pressure. In the gage pressure type, no evacuation is necessary. There are two
different ways to apply pressure to the diaphragm. With back side pressurization, as
illustrated in Figure 2. 7a, there is no interference with signal transducer, such as a
piezoresistor, that is normally implanted at the top surface of the diaphragm. The
other way of pressurization, i.e., front-side pressurization, Figure 2.7b, is used only
under very special circumstances because of the interference of the pressurizing
medium with the signal transducer. Signal transducers are rarely placed on the back
surface of the diaphragm because of space limitation as well as awkward access for
interconnects.
As shown in Figure 2. 7, the sensing element is usually made of thin silicon die
varying in size from a few micrometers to a few millimeters square. A cavity is cre-
ated from one side of the die by means of a microfabrication technique. The top sur-
face of the cavity forms the thin diaphragm that deforms under the applied pressure
from the measurand fluid. The thickness of the silicon diaphragm usually varies from
a few micrometer\ to tens of micrometers. A constraint ba\e made of metal (cal1ed a
header) or ceramic (Pyrex glass is a common material) supp<Jrts the silicon die The
deformati on of th e diaphra gm by the appli ed pressu re i) transduccd into electrical
signals by various lran'idu ction techniques, as wi ll be de)cribcd later in this section.
The assembly of the sensing clement s as shown in r◄ ig urt 2 7, together with the sig-
nal transduction element is th en packaged into a robu st casing made of metal, ce-
ramic, or plastic with proper passivati on of the di e.

Figure 2. 7 I Cross sections of micro pressure sensors.

Mea~ urand
fluid inl et

Silicon die
with
diaphragm

Constraint
base

' - _ Measurand
fluid inlet
(a) Back side pressurized (b) Front side pressurized

Figure 2.8 schematically illustrates a packaged pressure sensor. The top view of
the silicon die shows four piezoresistors (R 1, R2 , R3, and R4) implanted beneath the
surface of the silicon die. These piezoresistors convert the stresses induced in the sil-
icon diaphragm by the applied pressure into a change of electrical resistance, which
is then converted into voltage output by a Wheatstone bridge circuit as shown in the
figure. The piezoresistors are essentiaily miniaturized semiconductor strain gages,
which can produce the change of electrical resistance induced by mechanical
stresses. In the case illustrated in Figure 2.8, the resistors R1 and R3 are elongated the
stresses induced by the applied pressure. Such elongation causes an increase of elec-
trical resistance in these resistors, whereas the resistors R2 and R4 experience the
opposite resistance change. These changes of resistance as induced by the applied
measurand pressure are measured from the Wheatstone bridge in the dynamic de-
flection operation mode as
R1 R3 )
Vo = Vin ( RI + R4 - R2 + R3 [2.1]

where V0 and Yin are ·respectively measured voltage and supplied voltage to the
Wheatstone bridge.
We will learn the detail characteristics of piezoresistive materials in Chapter 7.
~in wire bonds are used to transmit the voltage change from the piezoresistors
bndge through the two metal pads and the interconnect pins as shown in the c:oss-
~ectton view of the sketch. A packaged pressure sensor of this type is shown
m the left of Figure l 2. We will deal with the packaging of these sensors in detail in
 44

Fi gu re 2 . 8 I A tyo c31
mi cro pr0ssuft."' Sl' ,s, ,r ,1, :-ic' nl' !y
R1, R:, Ri, R4 = p!c 'll'f , ,1,1 ,,1,

Me tal pJd

Top view of silicon die +

Wi re bon d Piezores1s1ors

Cir cui t
Sil ico n
dia phr agm Meta.I
casing
Py rex gla ss atta ch
con str ain ing
bas e or metal
hea der

Int erc onn ect ______. Pas sag e for

pre ssu riz ed
medium

Cr oss sec tio n

Ch ap ter 11. M icr o pre ssu

re sensors wi th pie zo res ist
go od lin ea r rel ati on sh ip b~ ors ha ve hig h ga ins an d
tween the in -~ l~ e str ess es ex hi bi t a
put. Ho we ve r, it ha s~ m~ an d the_r~~istance ch an ge
Jor ~rawback: 1t 1s tem pe ou t-
Si gn al transduct10n m nn rat ur e se ns iti ve . .
cro pre ssu re se ns or s ca n
tiv ity an d ac cu rac y req uir vary, de pe nd ing on th e
ed from the sensor. Th ere se ns i-
tion me tho ds de ve lop ed are se ve ral oth er sig na l tra ns
for the se sen so rs. We wi ll du c-
Fi gu re 2.9 ill us tra tes a mi ill us tra te tw o of th em he
cro pre ssu re se ns ing un it re.
for pr es su re me as ur em en uti liz ing ca pa cit an ce ch
t. Tw o ele ctr od es ma ~e an ge
the bo tto m of the top co ve of thi n me tal fil ms are att
r and the top of the dia ph ac he d to
ap hr ag m du e to the ap pli rag m. An y de fo rm ati on
ed ~r ess ure will na rro w the of the di-
lea din g to a ch an ge of ca ga p be tw ee n the tw o ele
pa cit an ce ac ros s the ele ctr ctr od es ,
tag e of be ing rel ati ve ly ind od es . Th is me tho d ha s the
ep en de nt of the op era tin g ad va n-
tem pe rat ur e. Th e ca pa cn
an ce C
 p cJtor can be rel t d to th Pd b t\\ n t p
fl
JOfl A
C - E,€o d
. (l,lJ
· . . o f the dJC Iectnc medrnm and
. the relative penrutt1v1ty PellnJ1.
Ssp ace (vacuu m); €o- = 8 85 pF/m (pF - picofarad - 10~,;sfthe
arad)
I~ ~Jucfhf.~:
uvJ l} o r
s us ng capac lance
fig ur e 2.9 t Micro pressure sensor
signal transduction.
Metallic
electrode Stlicon co\ er
/ Metallic
.--. ...- --t --- .... ..-- ::1 1~ --- .,
electrode

Silicon d1e

Constraint
base

~ ~1easuran d
fluid inlet

stance s is given in TabJe 2.2.

The relative pennittivity for several sub
e, of
Table 2.2 I Relative permittivity,
selected dielectric materials

Ma ter ial s Re lat ive pe rm itti vit y

Air 10
Paper 2-3.5
Porcelain 6-7
Mica 3-7
Transformer oil 4.5
Water 80
Silicon 12
Pyrex 47

s have ident1·
Determine the ca . plate d e/ectnc
. pac,tance of a parallel -pi ate capacit or Th e twoAir ,s the
cal dime .
ns,ons of L == W -- 1000 µm with a ga p d = 2 µm .
mediu m bet w
,~en the two plates .
:...~......,_,e
on
 ~ es gn and Manu•ac
ture

Capacitors ar~ co nu no
. . . . n trJ.nsduc~rs as \\ ell
capacn::mce \'ana as ac tm t rs · , _
. • tio n m J. c,rpac1tor can be meJsu ' 0 , m m11..:
illustrated m red b'- sm,nl ros\ "lems The
Figure 2.1 0 (Br.idle) .) ,
1- e 1.:ucu1t...
·
et al. 19011 such as tlnt
· '
Fi gu re 2. 10 I A typ 1
ca b'1dge for
1
capacitance measu1em
ents

The electrical bridge in

Figure 2.10 is similar to
ure 2.8 for resistance m the Wheatstone bridge
easurements. The varia in Fig-
measuring the output vo ble capacitance can be
ltage \~ The following rel me asured by
the variable capacitance ationship can be used to
in the circuit: determine

V,
ilC
0
= V
2(2C + ~C ) m [2.31
where AC is the capacit
ance change in the capa
C is the capacitance of citor in the micro pressu
the other capacitors in the re sensor and
constant supply voltage bridge. The bridge is su
Vin. bjected to a
Figure 2.11 illustrates
a manifold absolute pres
mobile [Chiou 1999). Th sure (M AP ) sensor for
is sensor uses a capacitor an auto-
arrangement of the silico as signal transducer. The de
n die in the pressure ch tailed
Figure 1.22c. amber is similar to tha
t shown in
Figure 2.11 I A manif
old absolute pre ssure sensor with capa
citor transducer

adhesl\'e

Well sealant
(Courtesy of Motoro
la Co rpo rat ion )
 CHAPTER 2 Working Principles of Microsystems

. sensors with capacitance signal transduction are not nearly as

Micro pressure .
. th applied pressure as those with piezoresistors . However, capacitance
ens1t1ve to e . h . .
s t early as sensitive to operating temperature as t e p1ezores1stors.
transducers are no n . .
h d one should be aware of a maJor shortcommg of capacitance trans-
On the ot her an ' . . . I' t th
. Th voltage output from a capacitance bridge ctrcmt is non mear o e
ducers. . • This
e acitance due to pressure vanat10n. ·
· non 1mear · tJ t t ela
mpu ou pu r -
Of Cap ·11 ·11 h
chancre
. e. • · d'cated in Equation (2.3). The following example w1 1 ustrate sue
uonsh1p 1s m 1
nonlinear relationship. Ij:f j?, j;
·ne the voltage output of the capacitance bridge circuit illustrated in Fig-
De terrn1 ·b d ·
with variation of the gap between two flat-plate electrodes as descn e in
ure 2.10
Example 2.1.

■ Solution •
we may use Equations (2.2) and (2.3) to establish the following table for voltage
output:
Capacitance C, Change of VoHage ratio
Gap d between
electrodes, µ.m pF capacitance 6.C, pF v.,v...
2.00 4.430
.. 0 0
1.75 5.063 0.633 0.033
1.50 5.910 1.477 0.071
1.00 8,860 4.430 0.167
0.75 11.813 7.383 0.227
.0.50 17.720 13.290 .::... 0.300

Figure 2.12a illustrates the construction of a nucro pressure sensor using a vi-
brating beam for signal transduction [Petersen et al. 1991]. A thin n-type silicon
beam is installed across a shallow cavity at the top surface of the silicon die. A p-type
electrode is diffused at the surface of that cavity under the beam. The p- and n-type
silicon layers are doped with, respectively boron and phosphorus, as will be de-
scribed in detail in Chapter 3. Both p- and n-type silicon are electrically conductive.
The beam is made to vibrate at its resonant frequency by applying an ac signal to the
diffused electrode in the beam before the app]ication of the pressure to the di-
aphragm. The stress induced in the diaphragm (and the die) will be transnutted to the
vibrating beam. The induced stress along the beam causes a shift of the resonant fre-
quency of the bec!m. The shift of the resonant frequency of the beam can be corre-
lated to the induced stress and thus to the pressure applied to the silicon diaphragm
Forrnul~s for calculating the corresponding resonant frequency shift in a vibrating
beam will be presented in Chapter 4.
Thi~ type of signal transduction is insensitive to temperature and provides ex-
cellent lmear output s·igna Is. Th e ct·1sadvantage of such sensors 1s
. the cost to fabricate
th em.
 M EM S an d M i
cr os ys te m s: De si gn
an d M an uf ac tu re
F ig u re 2. 1 2 I P re
ss ur e se ns or . g a vi.br .
s us in at in g be am si gn al tra
ns du ce r.
D iff us ed p- ty pe
. . Vi br at in g be am
S I 1IC OO di ap hr ag m el ec tro de n- ty pe Si wafer, 40 µm
wide 12 2 -- -- -- -- -- -- --
( 10 0 µm th ic k) X 60 0 µm Jong X -- -- --
6 µm th ic k -- -- -- -- --
=
N

Si lic on di e ✓ .:it:
..;
V
12 00 µm sq ua re C
~

X 40 0 µm th ic k =
Q"
t
"'-
108

20 120
Pr es su riz ed m ed iu m Co ns tra in t ba se Pr es su re , kP a
(a ) D ia gr am o f th e pr
es su re -s en si ng el em en t (b ) R ep or te d se ns iti
vi ty
 2. 5 I M IC R O A C C E L E R O M E T
ERS
An accelerometer is an instrument that
measures the acceleration (or decelerati
a moving solid. Microaccelerometers on) of
are used to detect the associated dy
forces in a mechanical system in motio namic
n. These accelerometers are widely use
automotive industry as described in Chap d in the
ter 1. For example, acceleration sensors
+2g range are used in a car's suspension in the
system and antilock braking system (A
whereas +SOg range acceleration sensor BS),
s are used to actuate air bags for drive
passenger safety in event of collision wi r and
th another vehicle or obstacles. The no
g represents the gravitational acce tation
leration, with a numerical value of
9.81 m/s 2. We present microaccelerom 32 ft/s 2 or
eter in a separate section because this
device is often classified as an inertia type of
sensor, yet it contains actuation eleme
Most accelerometers are built on the pri nts .
nciples of mechanical vibration , as wi
be described in detail in Chapter 4. Pri ll
ncipal components of an acceleromete
mass supported by springs. The mass is r are a
often attached to a dashpot that provid
necessary damping effect. The spring an es the
d the dashpot are in tum attached to a ca
as illustrated in Figure 2.32. sing,
In the case of micro accelerometers7
significantly different arrangements
necessat-y because of the very limited are
space available in microdevices . A mi
silicon beam with an attached mass nute
(often called a seismic mass) constitu
spring-mass system, and the air in the tes a
surrounding space is used to produce
damping effect. The structure that sup the
ports the mass acts as the spring. A
microaccelerometer is illustrated in Fig typical
ure 2.33; in it, the mass is attached to
tilever beam or plate, which is used as a can-
a spring. A piezoresistor is implanted
on the

C
 CH APTER 2 Working Principles of Microsystems

figure 2.32 I Typical arrangement of an accelerometer

The accelerometer ts
attached to the vibrating
solid body

Sprmg
k

Mass
M Dashpot
with
damping
Vibrating C
solid body

il
beam or plate to measure the defonnation of the attached mass, from which the am-
plitudes and thus the acceleration of the vibrating mass can be correlated. A mathe-
matical fonnula that relates the vibrating mass to the acceleration of the casing is
available in Chapter 4. Since acceleration (or deceleration) is related to the driving
dynamic force that causes the vibration of the solid body to which the casing is at-
tached, accurate measurement of acceleration can thus enable engineers to measure
the applied dynalT'ic force. It is not surprising to find that microaccelerometers are
widely used as a trigger to activate airbags in automobiles in an event of collision, and
also to sense the excessive vibration of the chassis of a vehicle from its suspension
system.

Figure 2.33 I Schematic structure of a microaccelerometer.

Silicon cantilever
beam or plate
~

Vibrating base

There are many different types of accelerometers available commercially. Signal

~~sduccrs used in mkroaccelerometers include piezoelectric, piezoresistive, capac-
itive, and r~sonant members [Madou 19971. Principles of signal conversion of these
schemes will be described in Chapter 11 .
 lr 'Ll Mr< rn:;y• lt•r11•, f)i • ,llJll lrH I M ' 1111 rf,1r fiJr<' J
Ml "'" • ,, sem.or) Jn t 1C L Ur
( rncrtia
• 1111uoat tclrrnmclN or. r ha deployment I~ au
P,., h.1p, ti"' '110,1 w 'dl'i y "" d .to " " I, tome! er Ior ,11 • '~c gra Ieel wllh st gna!
,.-111 m.11k,·1pl:1<·,· is lh,• "'.h"g1.'.1,•d ;:';,'.,d ·, 8. lhcsc devices ar c ' clement, i.e .. the ac
ton}L)hilc:-:. \s shown Ill Jiigu1u, J . cm.: u,t s. Tile.; \c.;n~mg 2 34 The working
11 1nsductwn ,md llic .,ssol·1' ill·d ckctwn1c r' m as illustrate d in Figure . .
l'l:Jt,ro,11L'ter. h,ls a spcnal ro11IJg11J1 ,1 l(n1ct••:1 is presented below.
pnnnpk ol this typt' o f nm.n')'ll'CC ero I,, •

Figure 2.34 I Schema1,c atrangcmcnl o f ,,, m,·c,o ,nortia sensor. Applied

Spnngs
. - acceleration
11,•1h,.-s)~

Proof mnss
(thin beam)
Proof mass --+

Different.Jal capacitive
sensor cells
■ : Anchors
Output
capacitance
changes
(a) In equilibrium condition
(Doscher (1999]) (b) With an acceleration toward the left

With reference to Figure 2.34a, a thin beam is attached to two tethers at both
ends. The tethers are made of elastic material and are anchored at one side as shown
in the figure. The thin beam acts as the seismic mass (called the proof mass) With an
electrode plate attached. The electrode plate that is attached to the proof mass is
placed between two fixed electrodes. In the event of an acceleration of the unit, the
proof mass will displace in the direction opposite to the acceleration, as shown in
Figure 2.34b. The movement of the proof mass induced by the acceleration (or de-
celeration) can be correlated with the capacitance change between the pair of the
electrodes.

We realize that the proof mass moves in the direction opposite to the accelera-
tion or deceleration of the unit. The arrangement in Figure 2.34 Will measure tl .
· · h d' ·
celeratwn only 10 t e rrectwn along the length of the proof mass. The arr 1e ac
. F. 2 3c:. d . d f h angement
10 . 1gure
uruts . ., 1s es,gne
&re attached to the same e measurements in both x and y drrections When both
or tbase.

A more compact arrangement is illustrated in Figure 2 36 · h" h

mass 1s replaced by a square plate that can displace b th .
.
r t O
Electromechanical design principles are presented in 10
, 1n w 1c the proof
a nd Y dtrec1tons. ~
Chau et al . 1995]. e erences !Doscher 1999,
 CH A P T E A 2 Working Princ1ploG of Micro y Jtnm

figure 2.35 I Schemati c d iag ram of a dunl-ax1al-mot1on ~,nr,•,or

■ Anchorn

Tether
I Cl)

~
e
t.-,
0
J>, oof w:p 1-,

0
....

\ ~~
Q.,

/
Differential capaciti ve
sensor cells
(y direction l) (x directlo,, < ►)

(Doscher (1999] .)

Figure 2.36 I Schematic diagram of a compact

dual-axial-motion sensor.

DifferentiaJ
capacitive
sensor cell
II
~
Tether------- .

■ Anchor
II
(Courtesy of Analog Devices Inc , Norwood, Massachusetts. )
 ~ --.~1--
_...-----L.X..\..CWJQ
_L:jlf--~~~~~~::..=_:
___~~~~~~~~____,j~
o~4-~~
~ .Tl ~-J~

'C V~ 7- -_
,_~>~~➔L
J -- - - ~ ~ ~ - - - - - : , -
- - - - -- - -~ ~ ~ 4 -_ _
"YVl
~ '"
tr <
I •