3–346
Motorola Sensor Device Data
For More Information On This Product,
Go to: www.freescale.com
!
Prepared by: Eric Jacobsen and Jeff Baum
Sensor Systems Engineering Group
Motorola Sensor Products Division
Phoenix, AZ
INTRODUCTION
When designing a circuit for a sensor system, it is desirable
to use fixed–value components in the design. This makes the
system easier and cheaper to produce in high volume. The
alternatives to using fixed–value circuitry are very expensive
and usually impractical: laser–trimming resistances, manually
calibrating potentiometers, or measuring and selecting
specific component values are all very labor–intensive
processes. However, every sensor has device–to–device
variations in offset output voltage, full–scale output voltage,
dynamic output voltage range (difference between the
full–scale output voltage and zero–scale output voltage which
is commonly referred to as the span), etc. Moreover, these
same parameters also vary with temperature — e.g.,
temperature coefficient of offset (TCVoff) and temperature
coefficient of full–scale span (TCVFSS). To further complicate
this situation, the fixed–value circuit in which a sensor is
applied also has variation — e.g., the voltage or current
regulator and resistors all have a specified tolerance.
Since today’s unamplified solid–state sensors typically
have an output voltage on the order of tens of millivolts
(Motorola’s basic 10 kPa pressure sensor, MPX10, has a
typical full–scale span of 58 mV, when powered with a 5 V
supply), a major part of the fixed–value circuitry is a gain stage
that amplifies the signal to a level that is large enough for
additional processing. Typically, this additional processing is
digitization of the amplified analog sensor signal by a
microcontroller’s A/D converter. To obtain the best signal
resolution with an A/D, the sensor’s amplified dynamic output
voltage range should fill as much of the A/D window (difference
between the A/D’s high and low reference voltages) as
possible without extending beyond the high and low reference
voltages (i.e., the zero–pressure offset voltage must be
greater than or equal to the low reference voltage, and the
full–scale output voltage must be less than or equal to the high
reference voltage). In any case, the device–to–device,
temperature, and circuit variations create a design dilemma:
with a fixed–value amplifier circuit, the gain as well as any dc
level shift incorporated in the amplifier design are fixed. If the
variation of any of the aforementioned sensor parameters is
too large, the amplified sensor output may saturate the
amplifier near either its high or low supply rail or may extend
beyond either the high or low reference voltages of the A/D
converter. In either case, error (non–linearity) results in the
system. To avoid this scenario, the solution is to design a
fixed–value circuit that optimizes performance (signal
resolution) while taking into account all possible types of
variation that may cause the sensor output to vary. In other
words, the goal of this fixed–value sensor system is to attain
the best performance possible while ensuring through design,
regardless of any system variation, that the sensor’s amplified
output will ALWAYS be within the saturation levels of the
amplifier and the high and low reference voltages of an A/D
converter.
The implication of ensuring that the sensor’s amplified
output is always unsaturated and within the high and low
reference voltages of the A/D is that an accurate software
calibration of the sensor’s output is possible. By sampling the
sensor’s output voltage at a couple of points at room
temperature (zero and full–scale output, for example), all the
room temperature device–to–device and circuit variations are
nullified. Obviously, temperature variations will create error in
the system (sensor’s output voltage will drift with changing
temperature), but, by design, the sensor’s output voltage will
remain within the A/D’s valid range.
This paper discusses a methodology that optimizes a
sensor
system’s
performance
device–to–device, temperature, and circuit variations that can
create variation in the amplified sensor output. The
methodology starts with a desired performance and some
established parameters and then considers each type of
variation in a worst case analysis to determine if the desired
performance is attainable. While this paper discusses this
methodology for pressure sensors and a specific amplifier
topology, the methodology is applicable to low–level,
differential–voltage output sensors and amplifier circuits in
general. Two specific examples are presented that apply this
methodology. The first example uses Motorola’s MPX10
pressure sensor, and the second example uses Motorola’s
MPX2010 pressure sensor. Both sensors have a full–scale
rated pressure of 10 kPa; the difference between the devices
is the MPX2010 has on–chip calibration and temperature
compensation circuitry to calibrate and temperature
compensate the zero–pressure offset voltage and span. The
comparison of these two devices will emphasize how
dramatically device–to–device and temperature variations, if
not compensated, can affect a system’s overall performance.
while
considering
SEMICONDUCTOR APPLICATION NOTE
F
Freescale Semiconductor, Inc.
n
.