 |
Description  |
|
|
BACKGROUND OF THE INVENTION
The present invention relates to a blood pressure measuring instrument,
e.g. a sphygmomanometer and, more particularly, to a compensation circuit
for such a blood pressure measuring instrument for detecting the Korotkoff
sound without suffering variations in the piezo-electric properties.
In a conventional sphygmomanometer, there was provided a pressure sensor
for determining the blood pressure and developing an oscillation frequency
corresponding to the same. Piezo-electric elements of the pressure sensor
was utilized for converting an amount of the blood pressure into the
oscillation frequency. However, the piezo-electric elements inevitably
suffer variations and nonlinear properties. This necessarily requires
accurate examination and modifications in the pressure sensor.
OBJECTS AND SUMMARY OF THE INVENTION
Accordingly, it is the primary object of the present invention to provide a
highly reliable sphygmomanometer.
It is another object of the present invention to provide an improved
sphygmomanometer where the necessity of examination and modification
procedures are completely eliminated.
It is still another object of the present invention to provide an improved
sphygmomanometer whose piezo-electric properties are controlled to
compensate for variations and nonlinear properties by electronic
techniques.
Other objects and further scope of applicability of the present invention
will become apparent from the detailed description given hereinafter. It
should be understood, however, that the detailed description and specific
examples, while indicating preferred embodiments of the invention, are
given by way of illustration only, since various changes and modifications
within the spirit and scope of the invention will become apparent to those
skilled in the art from this detailed description.
To achieve the above objects, pursuant to an embodiment of the present
invention, a sphygmomanometer comprises a variation compensation and a
lineality compensation circuit. The variation compensation circuit is
provided for cancelling variations in piezo-electric properties of a
pressure sensor equipped with a sphygmomanometer. The lineality
compensation circuit is utilized for compensating for nonlinear properties
in the piezo-electric properties of the same.
BRIEF DESCRIPTION OF THE DRAWINGS
The present invention will become more fully understood from the detailed
description given hereinbelow and the accompanying drawings which are
given by way of illustration only, and thus are not limitative of the
present invention and wherein:
FIG. 1 is a perspective view of a pressure sensor showing a principle of
the same;
FIG. 2 is a diagram of output properties of the pressure sensor illustrated
in FIG. 1;
FIG. 3 is a diagram of movement properties of a bellows included within the
pressure sensor;
FIG. 4 is a diagram of oscillation properties of the pressure sensor;
FIG. 5 is a block diagram of a control circuit of a sphygmomanometer
according to the present invention;
FIG. 6 is a time chart of various signals occurring in the control circuit
shown in FIG. 5;
FIG. 7 is a diagram of conversion properties of the pressure sensor showing
a model of the conversion properties;
FIG. 8 is a diagram of a polygonal line approximation function for the
pressure sensor;
FIG. 9 is a flow chart showing calculation operations;
FIG. 10 is a graph of a model of experimental measurements by the pressure
sensor;
FIG. 11 is a graph showing compensation principles for the variations of
the pressure sensor;
FIG. 12 is a graph showing reference polygonal line approximation functions
and correction amounts;
FIG. 13 is a circuit configuration of a variation compensation circuit
according to the present invention;
FIG. 14 is a time chart of various signals occurring in the variation
compensation circuit shown in FIG. 13;
FIG. 15 is a circuit configuration of a lineality compensation circuit
according to the present invention; and
FIG. 16 is a time chart of various signals occurring in the lineality
compensation circuit illustrated in FIG. 15.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
FIG. 1 shows a pressure sensor adapted to the present invention, FIG. 2 is
a graph of output characteristics of the pressure sensor shown in FIG. 1,
FIG. 3 is a graph of movement characteristics of a bellows employed within
the pressure sensor, and FIG. 4 is a diagram of frequency properties of
the pressure sensor.
With reference to FIG. 1, the pressure sensor comprises a bellows 1, a core
2, a coil 3, and an oscillation circuit 4. The bellows 1 lengthens and
shrinks responsive to applied pressure. The core 2 is disposed at the tip
of the bellows 1. The core 2 is removed within a cavity portion of the
coil 3. The inductance of the coil 3 is varied in accordance with the
displacement of the core 2 to change oscillation frequency developed from
the oscillation circuit 4, according to movement of the bellows 1.
The data in FIG. 2 are plotted with oscillation frequency on ordinate and
the length .DELTA..chi. of a portion of the core 2 inserted into the coil
3 on abscissa. L is the length of the core 2 and l is the total length of
the coil 3.
The value of the oscillation frequency f is a minimum as represented by
f.sub.L when .DELTA..chi.=(L+l)/2 because the inductance of the coil 3 is
a maximum.
The data in FIG. 2, in general, has a parabolic curvature with an axis of
.DELTA..chi.=(L+l)/2 and slightly linear characteristics in two portions
depicted in the regions I and II.
FIG. 3 shows a graph showing typical found values in the movement of the
bellows 1 according to the applied pressure between zero and 300 mm Hg. As
is viewed in FIG. 3 there are variations of the found values in the
bellows 1 between .+-.10% by manufacturing conditions for the bellows 1.
However, these variations allow the slope of the characteristic graph to
only change. In other words, no two characteristics lines cross each
other.
It will be apparent from FIGS. 2 and 3 that the movement of the bellows 1
is saturated according to the increase in the applied pressure and the
change in the oscillation frequency should be referred to in the section
of FIG. 2 where the change in the oscillation frequency of the bellows
changes according to the variation in the length of the overlap
.DELTA..chi..
FIG. 4 represents by solid line oscillation properties of the pressure
sensor which is manufactured for the test under the above-mentioned
condition. As is apparent from FIG. 4, the error is estimated to be the
worst within .+-.1 mmHg even if a line is utilized for making an
approximation at the section defined by the pressure between zero to 100
mmHg. In connection with the pressure above 100 mmHg, the oscillation
properties show slightly saturated conditions which cause the bellows to
fail to hold completely its lineality properties. The dotted lines of FIG.
4 show other undesirable characteristic properties with different slopes
due to the variations in the manufacturing conditions and/or the
variations in the location of the coil 3 and the core 2.
Now with reference to FIG. 5, there is shown a block diagram of a control
circuit of a digital sphygmomanometer according to the present invention.
The control circuit comprises a pressure converter 5 referred to the
pressure sensor, a counter 7, a variation compensation circuit 8, a
lineality compensation circuit 9, a determination circuit 10, a clock
circuit 11, a divider 12, an amplifier 13, a processor 14, a driver 15,
and a display 16. The output of the pressure converter 5 is applied to the
counter 7 through a gate 6. The variation compensation circuit 8 functions
to generate polygonal line approximation functions corresponding to the
pressures detected by the pressure sensor by modifying a reference
polygonal line approximation function according to modification codes
applied thereto.
The lineality compensation circuit 9 serves to calculate pressure values
corresponding to the modified polygonal line approximation functions by
information detected by the pressure sensor for a predetermined time
period. The determination circuit 10 is operated to determine both the
increase and the decrease of the pressure and whether or not the present
pressure is 20 mmHg or more, thereby providing respective control signals.
The amplifier 13 operates to amplify and normalize the Korotkoff sounds
sensed by a microphone (not shown) to thereby provide the output impressed
onto the processor 14. The processor 14 is provided for picking up the
true Korotkoff sounds and applying them to the driver 15. The display 16
indicates systolic and diastolic pressures responsive to the output of the
driver 15.
The operation of the control circuit is described with reference to a time
chart shown in FIG. 6. Assumed that the modification codes are preset to
be suitable for the pressure sensor equipped within the system, a start
signal is applied to the variation correction circuit 8 in response to the
energization of the power supply. The variation compensation circuit 8
sets a polygonal line approximation function according to the modification
codes. The pressure converter 5 provides different oscillation frequencies
depending on the relative pressure.
The clock circuit 11 develops clock signals 17 having a small duty factor
which functions to make the gate 6 conductive. The counter 7 counts the
oscillation frequency. The lineality compensation circuit 9 makes
lineality compensation according to a timing signal 22 after the counting
is completed. The data is transferred into the determination circuit 10
and the driver 15. The determination circuit 10 determines whether the
relative pressure is 20 mmHg or more. The determination circuit 10 further
compares the now received data to the previously present data which is
sensed in the previous timing, using the comparison timing developed by
the divider 12. The system thereby senses an increase or decrease in the
pressure. When the pressure data is below 20 mmHg, the determination
circuit 10 develops reset signals for the driver 15. If the pressure data
is determined to be decreasing as compared to the same in the preceding
time period, pressure decrease recognition signals generated from the
determination circuit 10 are impressed into the processor 14. The
processor 14 recognizes the signals developed from the amplifier 13 to
determine if they are true Korotkoff sounds using the timer means
contained therein and the signals derived from the determination circuit
10. The results by the processor 14 are transferred into the driver 15.
The driver 15 controls the timing where the data derived from the
lineality compensation circuit 9 is indicated in the display 16, depending
on the data developed from the determination circuit 10 and the processor
14.
Blood pressure measuring procedures are traced according to pressure change
10 of FIG. 6 as follows. While there is no pressure, a first indicator I
referred to an indicator for showing systolic pressure is blank condition
20. A second indicator II normally displays the pressure data which is now
measured and it indicates diastolic pressure when the diastolic pressure
is detected. The second indicator II displays zero pressure or the
diastolic pressure in the preceding measuring procedure in the timing 21
while the measured pressure is less than 20 mmHg.
While the measured pressure is 20 mmHg or more, the pressure data
determined by the pressure sensor is lineality compensated using lineality
compensation timing signals 22 and the results are transferred, for
displaying purposes, into the second indicator II each time the
calculation is completed. The decrease of the pressure data is recognized
by comparing the pressure data introduced into the determination circuit
10 at a timing 24 with the same at the preceding timing 25. Upon the
detection of the decrease of the pressure data, the pressure decrease
recognition signals are introduced from the determination circuit to the
processor 14.
Since the microphone for collecting the Korotkoff sound inevitably detects
various noise, the output of the amplifier 13 necessarily contains noise
signals except for signals in synchronization with pulsation as viewed in
amplifier output signals 26 of FIG. 6. Therefore, the true Korotkoff sound
is obtained by executing the AND logic operation on the pressure decrease
recognition signals and the amplifier output signals 26 and further
eliminating, using the timer means, noise signals contained in pressure
decreasing time periods. Processor output signals 27 are true Korotkoff
sound.
The driver 15 allows the first indicator I to indicate and retain the
present pressure data as the systolic pressure, assuming that the first
signal 28 of the processor output signals 27 is the first Korotkoff sound.
The second indicator II, on the other hand, goes on refreshing the display
data in synchronization with the output signals of the processor 14. The
second indicator II displays the relevant pressure data as the diastolic
pressure only when the last signal 27 of the processor output signals 27
is regarded as the last Korotkoff sound. The determination whether a
further Korotkoff sound is present or absent is affirmed by not receiving
the output signals from the processor 14 with the elapse of a
predetermined time period of the generation of the last signal 29.
The variation compensation circuit 8 and the lineality compensation circuit
9 are now described in detail hereinbelow. First of all, the principle of
compensation utilized therein is explained. The principle is described
using methods of least squares with reference to a model of conversion
characteristics in the present sensor shown in FIG. 7. In these
statistical researches, the evaluation of errors in found values is
conducted after compensation functions are defined according to the given
data. However, the evaluation of the errors is not necessarily required
and, instead, the most suitable correlation functions are available within
a given range of errors and regions of two kinds of transformation are
determined which are available for the correlation functions.
Trial and error procedures by a computer are suitable for obtaining the
relevant correlation functions relied upon the given errors and,
simultaneously, defining the regions where thus obtained correlation
functions are available. At first, values of allowed errors are assumed in
accordance with the accuracy of the instruments. Thereafter, the regions
are desirably selected where the correlation functions are defined.
With reference to FIG. 7, values C1, 0 and C1, 1 of the found value X of
the pressure sensor are desirably selected and the found value X defined
by the region C0.ltoreq.X.ltoreq.C1, 0 is considered as below. A
correlation function (a linear function y=m x +b) is utilized for assuming
the data measured by the pressure sensor. Another correlation function is
defined by making difference values between the found values and the
linear function and adding least squaring of the respective difference
values each other and, at last, being selected to minimize the added
results. The evaluation of the error in the found values by the pressure
sensor is effected according to the obtained correlation function over the
region C0.ltoreq.X.ltoreq.C1, 0 to compare the measured error with the
allowed error.
Another value C1, 1 smaller than the value C1, 0 is selected when the
values of the measured error are more than the allowed error and vice
versa. The above-mentioned mathematical procedures are repeated until the
values of the measured error is equal to that of the allowed error. This
repetition is provided within the given allowed error both the correlation
functions having the widest variation region and the measuring region
available by the pressure sensor.
Referring now to FIG. 7, this means that a plurality of coefficients C1,
A1, and B1 are defined in C0.ltoreq.X.ltoreq.C1 and P=A1X+B1. A plurality
of equations P=A2X+B2 . . . P=ANX+BN are determined in the respective
sections by repeating the preceding procedures in the sections
C1.ltoreq.X.ltoreq.C2, C2.ltoreq.X.ltoreq.C3 . . . ,
CN-1.ltoreq.X.ltoreq.CN. It is preferable in obtaining the correlation
function P=A2X+B2 in the section C1.ltoreq.X.ltoreq.C2 that the
correlation function P=A2X+B2 be continuous at X=C1 with the correlation
function P=A1X+B1. The condition of accomplishing the continuous
connection at X=C1 is B2=(A1-A2)C1+B1, e.g. to require defining the values
A2 and C2.
The repetition of the above-mentioned procedures over the total sections of
measuring procedures by the pressure sensor provides a series of polygonal
line correlation functions which are referred to conversion characterizing
curves within the allowed error for converting from the pressure by the
pressure sensor to electrical signals. The polygonal line correlation
functions are linear functions and continuous with each other. These are
referred to polygonal line approximation functions. At a plurality of
sectional points DK (K=0, 1, 2, . . N), the polygonal line correlation
functions show different slopes. The values of the pressure at the
respective sectional points are derived from the equations P.sub.0
=A1Co+B1, PK=AK CK+BK (K=1, 2, . . . , N). The respective polygonal line
correlation functions are featured by the pressure values PK at the
respective sectional points and the found values CK (converted value) by
the pressure sensor corresponding to the pressure value PK.
Using the thus derived polygonal line approximation functions, the pressure
values are determined by the found values measured by the pressure sensor
as described below. It is assumed that the polygonal line approximation
functions as illustrated in FIG. 8 are obtained under the data PK and CK
at the respective sectional points (K=0, 1, 2, . . . , N). A found value
CM is measured by the pressure sensor in assumptions. The logical
principle for changing from the found value CM to the pressure value is
assumed as follows.
The data CK and PK at the respective sectional points are subsequently
stored in a memory circuit. The comparison between the found value CM and
the data CK (K=0, 1, . . . , N) at the respective sectional points is
conducted within a comparator to determine which sectional regions contain
the found value CM. The following calculation is carried out at the
detected sectional region for the found value CM.
##EQU1##
The above two equations are mathematically equivalent to each other.
However, this contains subtill differences in the manner of calculating
PM. These differences can be eliminated by changing detection methods for
determining which sectional regions include the found value CM.
A flow chart of operating the calculation according to the equation (2) is
depicted in FIG. 9. Four registers A, B, C, and D are utilized for
conducting the calculation in the following order.
1 The found value CM by the pressure sensor is transferred into the B
register so that address counters of the memory circuit are allowed to be
reset which store the data at the sectional points in the conversion
characterizing curve of the pressure sensor.
2 One of the measured values Ck is transferred into the A register, the
measured value Ck being selected by the address counter. The measured
values Ck is included within the conversion characterizing curve of the
pressure sensor in the respective sectional points.
3 The comparison between the data contained in the A register and the
same of the B register is carried out.
4 The data Ck and Pk selected by the address counter of the memory
circuit are transferred respectively to the C and D registers, when the
contents of the B register are more than that of the A register. The
address counter of the memory circuit is advanced by one. Therefore, the
procedure for transferring the data Ck into the A register is reproduced.
This repetition is carried out until the contents of the A register are
equivalent to or more than that of the B register.
5 When the contents of the A register are equivalent to or more than that
of B register, the contents of the B register are subtracted from that of
the A register and the results are retained in the A register.
6 The following calculation is carried out and the results are kept in
the B register.
##EQU2##
where the data Ck-1 and Pk-1 are stored respectively in the C and D
registers, and the data Ck and Pk are selected at present by the address
counter of the memory circuit.
7 The data Pk are introduced into the C register. The B register contains
values of the variation of the measured values by the pressure sensor
corresponding to an unit variation of the pressure. The unit of the data
Pk and Pk-1 is, for example, mmHg, the B register includes the amount of
the variation in the measured data by the pressure sensor corresponding to
the change in pressure of 1 mmHg. This means that the data ten times the
data contained in the B register are the amount of the variation in the
measured values relied upon the change of the pressure of 10 mmHg. For the
purposes of shortening calculation time periods, one unit reference
pressure is defined as follows while the calculation of
##EQU3##
is carried out. The one unit of the reference pressure is approximately
equivalent to the amount of the change in the pressure at the respective
sections when the available regions selected from the conversion
characterizing curve of the pressure sensor are divided into the number N.
The D register is allowed to store the reference pressure (one unit of the
pressure).
8 The above-mentioned data is completed and the comparison between the
contents of the A register and that of the B register takes place.
9 When the contents of the A register are equivalent to or more than that
of the B register, the contents of the B register are subtracted from that
of the A register and the results are kept in the A register. The contents
of the D register are subtracted from that of the C register and the
results are stored in the C register. Therefore, the comparison between
the contents of the A and B registers is reproduced. This cycles are
repeated before the contents of the B register become more than equal to
that of the A register.
10 When the above conditions are completed, shift operations are carried
out in the B and D registers to reduce the order of the contents therein.
11 The completion of the calculation procedures is determined by either
making completely setting of zero conditions in the D register or reducing
the contents of the D register by one at each shifting procedure with a
predetermined number of occurrence of the shifting procedures. If the
completion does not reach, the comparison between the contents of the A
and B registers are still carried out before the shifting procedures are
terminated. The results of the above-mentioned calculation procedures are
stored in the C register.
The following is an example for setting an unit of the pressure. It is rare
that the divided sections of the convention characterizing curve of the
pressure sensor have equivalent intervals each other. However, it is
assumed for convenience that the variation in the pressure at one of the
sections is about 80 mmHg. The unit of the pressure is set to be several
tens mmHg as described previously. The unit of the pressure is available
in 10, 20, 30 mmHg or the like. The amounts of the variation in the data
measured by the pressure sensor per the variation of an unit of the
pressure (1 mmHg) are calculated by the following equation.
##EQU4##
The B register receives the amounts ten or twenty times the
above-determined values. 10 mmHg or 20 mmHg are set for the D register.
The principle for compensating for the variation in the pressure sensor is
described hereinbelow. FIG. 10 shows in model manners tendency of the
variations in the found values of the pressure sensor where a plurality of
characterizing curves a to f exist in general and a characterizing curve g
is rare and another characterizing curve h is not in practice measured.
Logical correction is, therefore, applied to the characterizing curves a
to f not to g and h.
In FIG. 10, it is assumed that when the different pressure P.alpha. and
P.beta. are measured by three kinds of pressure sensors a, b, and c, the
found values Ca.alpha., Ca.beta., Cb.alpha., Cb.beta., Cc.alpha. and
Cc.beta. are obtained. The following correlation is experimentally
confirmed between these found values.
##EQU5##
The correction principle is relied upon the equation (3). The following
equation (4) is obtained from the equation (3).
##EQU6##
Considering that the lineality compensation is carried out with reliance
upon the polygonal line approximation functions, it is preferable that the
values .alpha. and .beta. correspond to desired sectional points because
the values .alpha. and .beta. are arbitrary values. Using the equation
(4), the data Ca.alpha. and Ca.beta. at the sectional points can be
calculated for identifying the polygonal line approximation function if a
reference polygonal line approximation function C, a polygonal line
approximation function b on a desired characterizing curve and the
constant are determined. Therefore, a polygonal line approximation
function a is defined to determine an amount of (Cc.alpha.-Cb.alpha.) in
the equation (4). Now the amount of (Cc.alpha.-Cb.alpha.) is referred to a
reference correction value for determining purposes. The constant in the
equation (4) is selected to be effective only in its integral number in a
relationship with the allowed error. Code performance used herein means
that compensation amounts are converted into a constant using the
reference correction value and thus converted constant is termed codes.
Therefore, the compensation amounts can be represented using less kinds of
amounts to make broad compensation applications to be available. This can
be referred to a kind of the compression for the data.
FIG. 11 is utilized for explaining the reasonableness of the compensation
principle for the variation, where a polygonal line approximation function
1 is selected on the base of characteristics of a pressure sensor. A
measured value Cs is obtained by measuring a pressure Ps using another
pressure sensor. The measured value Cs is linearly approximated by the
polygonal line approximation function 1 to obtain the results P's
calculated by the lineality compensation circuit. However, the pressure Ps
can be assumed to be equivalent to the pressure P's if the following
inequality (5) is set up.
.vertline.Ps-P's.vertline.<allowed error (5)
The linearlity compensation on the base of the polygonal approximation
function 1 is finally available in the region 1 which is depicted by
dotted lines in FIG. 11. This means that the error of the lineality
compensation is within the allowed error regarding the various
characteristics included within the region 1 of FIG. 11 if the polygonal
line approximation function 1 is utilized for approximating purposes. If
the region 1 is not available, another region 2 is considered to define
another suitable polygonal line approximation function 2. It will be
apparent that the region 2 is preferably selected to be contact with the
region 1. The extent of the region 2 is, of course, determined according
to the inequality (5). The compensation of the variation in the
characteristics of the pressure sensor is widely possible by repeating the
above-mentioned procedures. To determine the regions, it is necessary to
define the respective polygonal line approximation functions
characterizing the regions.
As described previously concerning the definition of the code performance
and the reference compensation amounts, the polygonal line approximation
functions are determined with reliance upon the spaces of the reference
compensation amounts. Therefore, the respective polygonal line
approximation functions are defined according to the reference
compensation amounts so that the regions are fixed where the respective
polygonal approximation functions are effective.
There is a restriction regarding the determination of the reference
compensation amounts. The following is the fluctuation in the found values
by the pressure sensor in measuring the same pressure on the base of the
allowed error in the same polygonal line approximation function.
.+-.(the allowed error.times.the slope of the polygonal line approximation
function) (6)
On the other hand, the following relationship is valid.
The reference correction amounts>(the allowed error.times.the slope of the
polygonal line approximation function K)+(the allowed error.times.the
slope of the polygonal line approximation function K+1) (7)
The latter .perspectiveto.2.times.(the allowed error.times.the slope of the
polygonal line approximation function K) (8)
The approximation of the equation (8) is applied when the slope of the
polygonal line approximation function does not change. When the reference
compensation amounts are fixed under the conditions of the inequality (7),
there is unfortunately at least one region which is not corrected (has an
error more than the allowed error) to provide dissatisfactory results. The
reference compensation values should be selected to be valid in the
following inequality (9).
The reference correction values<(the allowed error.times.the slope of the
polygonal line approximation function K)+(the allowed error.times.the
slope of the polygonal line approximation function K+1) (9)
Two pressure sensor a and b which have different characteristics each other
measure the pressure to determine the polygonal line approximation
functions. The value .alpha. is defined according to the found values Ca
and Cb by the pressure sensor at the sectional points at which the maximum
of the compensation values is needed as follows.
(Ca-Cb)/.alpha.<2.times.(the allowed error.times.the slope of the polygonal
line approximation functions a or b) (10)
The reference correction value Mk [k=0, 1, 2, . . . , N (k is the number of
the sectional points)] is determined by the following equation (11).
##EQU7##
The standard for the compensation is required to introduce the principle.
The determination of the reference polygonal line approximation functions,
which are polygonal line approximation function as the standard for the
compensation, has a relationship with the effective utilization of the
compensation codes. It will be apparent that the selection of the
reference polygonal ling approximation function is effectively fixed by
either the characterizing curves of the pressure sensor emerged at the
ends of the region where the pressure sensor reveals its characteristics
or ordinary characterizing curve. With reference to FIG. 10, the utilized
characterizing curves are referred to a or f, otherwise, c or d.
FIG. 12 depicts in a mode manner one reference polygonal line approximation
function 0 and the reference compensation value Mk (k=0, 1, . . . , N) in
the respective sectional points assuming the found values of several tens.
The data of the reference polygonal line approximation function in the
respective sectional points is the found values Lk (k=0, 1, . . . , N) of
the pressure sensor concerning the pressure value Pk. To determine the
codes for one pressure sensor, the pressure is increased to suitable
sectional points. The value .alpha. is calculated by the following
equation.
##EQU8##
where Lz is the found value of the pressure sensor when the pressure is
increased to PN in FIG. 10.
The codes are selected in accordance with the value .alpha. as follows.
______ | | |
