MAX1450 Datasheet, PDF (9/12 Pages) Maxim Integrated Products – Low-Cost, 1%-Accurate Signal Conditioner for Piezoresistive Sensors

Low-Cost, 1%-Accurate Signal Conditioner for
Piezoresistive Sensors
OFFSET Calibration
Accomplish offset calibration by applying a voltage to the
OFFSET pin (SOFF determines the polarity of V_OFFSET).
This voltage is generated by a resistor-divider between
V_DD and V_SS (R_OFFA and R_OFFB in Figure 4). To calibrate
the offset, set V_OFFSET to 0 and perform a minimum pres-
sure input reading at room temperature. If the output volt-
age (V_OFFZERO) is greater than 0.5V, connect SOFF to
V_SS; if V_OFFZERO is less than 0.5V, connect SOFF to V_DD.
Adjust V_OFFSET until V_OUT = 0.5V.
Note that the magnitude of V_OFFSET is directly proportion-
al to the gain of the PGA. Therefore, if the PGA gain
changes after performing the offset calibration, the offset
must be recalibrated.
Linearity Calibration (optional)
Correct pressure linearity by using feedback from the
output voltage (V_OUT) to ISRC to modulate the current
source. If a bridge current is constant with applied
pressure, sensor linearity remains unaffected. If, with a
constant bridge current, the output voltage is nonlinear
with applied pressure (e.g., increasing faster than the
pressure), use pressure linearity correction to linearize
the output.
Performing linearity corrections through the use of a
transfer function is not practical, since a number of
required system variables cannot easily be measured
with a high enough degree of accuracy. Therefore, use a
simple empirical approach. Figure 5 shows the uncom-
pensated pressure linearity error of a silicon PRT. The
magnitude of this error is usually well below 1% of span.
Curves A, B, C, D, E, and F in Figure 5 represent increas-
ing amounts of linearity error corrections, corresponding
to decreasing values in the resistance of R_LIN. To correct
pressure linearity errors, use the following equation to
determine the appropriate range for R_LIN:
( ) ^{RLIN ≈}
2 R_ISRC x R_STC
R_ISRC + R_STC x S(p)
where S(p) is the sensitivity linearity error as % best
straight-line fit (BSLF). Ideally, this variable resistor
should be disconnected during temperature error com-
pensation. If this is not possible, set it to the maximum
available value.
First measure the magnitude of the uncorrected error
(R_LIN = maximum value), then choose an arbitrary
value for R_LIN (approximately 50% of maximum value).
Measuring the new linearity error establishes a linear
relationship between the amount of linearity correction
and the value of R_LIN.
Note that if pressure linearity correction is to be per-
formed, it must occur after temperature compensation
is completed. A minor readjustment to the FSO and
OFFSET will be required after linearity correction is per-
formed. If pressure linearity correction is not required,
remove R_LIN.
Ratiometric Output Configuration
Ratiometric output configuration provides an output that
is proportional to the power-supply voltage. When used
with ratiometric A/D converters, this output provides
digital pressure values independent of supply voltage.
Most automotive and some industrial applications
require ratiometric outputs.
The MAX1450 has been designed to provide a high-
performance ratiometric output with a minimum number
of external components (Figure 4).
Sensor Calibration and
Compensation Example
Calibration and compensation requirements for a sensor
involve conversion of a sensor-specific performance
into a normalized output curve. Table 4 shows an
example of the MAX1450’s capabilities.
A repeatable piezoresistive sensor with an initial offset
of 30mV and FSO of 37.5mV was converted into a com-
pensated transducer (using the piezoresistive sensor
with the MAX1450) with an offset of 0.5V and an FSO of
4.0V. The temperature errors, which were on the order
of -17% for the offset TC and -35% for the FSO TC, were
reduced to about ±1% FSO. The graphs of Figure 6
show the outputs of the uncompensated sensor and the
compensated transducer.
LINEARITY
ERROR
A
B
C
D
UNCOMPENSATED ERROR
(R_LIN REMOVED)
E
F
OVERCOMPENSATED ERROR
(R_LIN TOO SMALL)
PRESSURE
Figure 5. Effect of R_LIN on Linearity Corrections
_______________________________________________________________________________________ 9