lmp7731mfx National Semiconductor Corporation, lmp7731mfx Datasheet - Page 17
![no-image](/images/no-image-200.jpg)
lmp7731mfx
Manufacturer Part Number
lmp7731mfx
Description
2.9 Nv/sqrt Hz Low Noise, Precision, Rrio, Operational Amplifier In Sot23-5
Manufacturer
National Semiconductor Corporation
Datasheet
1.LMP7731MFX.pdf
(20 pages)
Thermopiles have very wide temperature ranges of -100°C to
1000°C
When choosing a thermopile for a certain application, one
must pay attention to several parameters.
Thermopiles' sensitivity, or responsivity, is determined by the
ratio of output voltage to the absorbed input signal power and
is usually specified in V/W. Typical sensitivity of thermopiles
ranges from 10s of V/W to about 100 V/W. Generally, higher
values of sensitivity are desirable. Sensitivity is dependent on
the absorber's area and number of thermocouples used in the
sensor. Sensitivity is often represented by S where:
S = V
The sensitivity of a thermopile varies with change in temper-
ature. This change is usually specified as the Temperature
Coefficient, TC, of sensitivity. Lower numbers are desired for
this parameter.
Resistance of the thermopiles is usually specified in the
datasheet. This is the impedance which will be seen by the
input of the amplifier used to process the thermopile's output
signal. Typical values for thermopile resistance, R
from 10s of kilo-ohms to about 100 kΩ. This resistance is also
a function of temperature. The temperature coefficient of the
resistance is usually specified in a thermopile's datasheet. As
with any other parameter, minimum variation with tempera-
ture is desired.
The dominant noise source for a thermopile is its resistance.
The noise spectral density of a resistor is calculated by:
Where k is the Boltzman constant and T is absolute temper-
ature. The unit of noise spectral density is: V/
For the thermopile sensor, this noise is usually represented
by V
A typical value for this voltage noise is in the order of a few
tens of nV/
The Noise Equivalent Power, NEP, is often used to specify
the minimum detectable signal level per square root band-
width. A smaller NEP is desired, however NEP is dependent
on the thermopile active area, A
NOISE
OUT
/P
where:
IN
.
FIGURE 7. Thermopile
D
. For a thermopile
TP
, range
20175207
17
and
As shown above, the NEP of two thermopiles cannot be com-
pared without considering the corresponding active areas.
A better way to compare thermopiles is to look at their specific
detectivity, D
noise and its sensitivity. It is normalized with respect to the
detector's active area and also noise bandwidth. D
by:
The unit of D
tivity range from 10
After receiving radiation, the thermopile takes some time be-
fore it comes to thermal equilibrium. The time it takes for the
sensor to achieve this equilibrium is called response time or
time constant of the sensor. Clearly, lower time constants are
very desirable.
Precision Amplifier
Since the output of thermopiles is usually very small and at
most in the order of only a few millivolts, the first part of the
signal conditioning path should involve amplification. In
choosing an amplifier for this purpose, a few different sensor
characteristics and the way they interface with the amplifier
should be considered.
Sensor's Impedance and Op Amp's Input Bias Current:
The input bias current causes a voltage drop across the sen-
sor and the amount of this voltage is equal to the sensor's
impedance multiplied by the magnitude of bias current. The
higher the sensor's input impedance, the more accentuated
the effect of the amplifier's input bias current will be. For very
high impedance sensors, it is imperative that op amps with
very low input bias currents be used. Thermopiles have input
impedances in the range of 100 kΩ, so input bias current is
not as critical as in some other applications.
Sensor's Output Voltage Range:
The output signal of the sensor is fed into the op amp where
it will be amplified or otherwise conditioned, e.g. level shifted,
buffered. It is important to pay attention to different parame-
ters of this output signal.
The lowest expected level of the sensor's output is very im-
portant. It is necessary to compare this level with the different
parameters contributing to the amplifier's total input noise. If
the sensor's output level is in the same order of magnitude or
smaller than the op amp's total input noise, then signal in-
tegrity at the op amp's output and the ADC's input will be
compromised.
*
*
. Specific detectivity includes both the device
is cm
8
to 3*10
/ W. Typical values for specific detec-
8
cm
/ W.
www.national.com
*
is given