There are many radiometric parameters for detectors, but responsivity,
noise-equivalent-power, detectivity and D* are the
most useful for the IR sensor designer. They can be both blackbody or spectral and are
defined as follows:
Responsivity, R:
The rms. signal voltage produced by unit rms. irradiance.
Noise-Equivalent-Power, NEP:
The rms. irradiance that produces unity signal-to-noise ratio in a specified bandwidth.
Unit is watts per root Hz.
Detectivity, D:
Reciprocal of NEP
D-star, D*:
A normalized detectivity with respect to bandwidth, BW, and detector area, Ad. A figure of
merit so that detectors with different sized elements can be compared.
Laboratory Blackbody Instrument
The radiometric parameters are arrived at by calculation from basic signals measured
with a laboratory blackbody. A blackbody is a heated target having near ideal radiation
characteristics. A perfect blackbody will radiate (or absorb) energy with 100% efficiency.
Thermal efficiency is described by a body's emissivity. An
emissivity of 1 is 100% efficient and is the highest value possible. Laboratory
blackbodies are about 0.98-0.99.
In order to make the blackbody visible to the detector, many of which are AC coupled
and adapt to their received radiation, it is equipped with a mechanical chopper and/or
shutter. The chopper can be run over a range of speed to measure the detector's frequency
response. Also needed is a measuring slit which is placed directly in in front of and very
close to the detector. It serves to direct radiation onto one element only, occluding the
other, to avoid errors introduced by common mode element connection.
Detector Parameters Measured with a Blackbody
Signal Output:
The voltage produced by the detector by the chopped incident irradiance. This
determines the gain necessary to interact with the chosen detection threshold.
Frequency Response:
The signal voltage at various chopper frequencies. Used to determine the target
velocity compensation for the processing amplifier.
Static Balance:
The percentage difference in signal voltage between detector elements.
Static Balance = 2.|A-B|/(A+B) X 100%, for a 2 element pyro.
where A & B are the elements.
This parameter can be measured with the chopper and slit and is the usual way that balance
is measured. It has proven not to be of significant importance
in actual application. This is because even though a perfect balance result might lead to
the conclusion that background cancellation would also be perfect, this may not be so
because of phase differences between the voltages produced by the elements. Thus although
the individual element voltages might be equal the phase shift produces an output which
may be significant. Phase differences can result from asymmetric heat losses in the
detector substrate and mounting structure.
Dynamic Balance:
This method of balance is measured by exposing the elements to blackbody
radiation by means of a shutter. First each element is exposed separately by turning the
radiation 'ON' and then 'OFF' after stabilization. The peak-to-peak voltage for each
element is recorded. Then both elements are exposed in the same way simultaneously and the
peak-to-peak voltage recorded.
Dynamic Balance = 2.ABptp/(Aptp+Bptp) X 100%, for a 2 element pyro.
where A & B are the elements.
This is exactly the way a detector will behave in an optical system.
Step Response:
This is the voltage output of the detector when exposed to radiation via a
shutter as described for dynamic balance. If the result is further analyzed by fast
Fourier transform, all significant performance parameters for both amplitude and frequency
can be determined. This is becoming the preferred method for serious application analysis.
Noise and Signal-to-noise ratio:
The noise output of the detector is measured as a peak-to-peak maximum over a
measuring period of usually many minutes. Signal-to-noise ratio compares the peak-to-peak
noise with the peak-to-peak signal measured with the blackbody. This parameter determines
the limitation in detector performance. From this result the number and complexity of
multiple optical fields-of-view can be determined. Working with marginal signal-to-noise
raises false alarm risks.
Return to top
FUJI &
CO.(Piezo Science)
