Noise equivalent power measurement

There are important figures of merit (5) that describe the performance of a photodetector. These are responsivity, noise, noise equivalent power, detectivity, and response time (2,6). However, there are several related parameters of measurement, eg, temperature of operation, bias power, spectral response, background photon flux, noise spectra, impedance, and linearity. Operational concerns include detector-element size, uniformity of response, array density, reflabiUty, cooling time, radiation tolerance, vibration and shock resistance, shelf life, availabiUty of arrays, and cost. 

The SNR of the detected signal is defined as the ratio of the signal change (produced as a result of the intensity modulation in the measurement cell) to the noise equivalent power (NEP) of the detection system for a given average received light intensity. In order to derive a figure for the NEP, various assumptions about the optical receiver must first be made. 

The responsivity (E) or specific detectivity (D ) and the noise equivalent power NEP (Wn), are often used to measure the sensitivity of a detector. The responsivity depends on the wavelength of the radiation and the temperature of the detector. The NEP, also called minimum detectable power, is the quotient of detector noise (N) divided by voltage responsivity (E). The D is the reciprocal of NEP, thus W = NIE and D = 1/Wn- A more sensitive detector has a smaller NEP and larger D, which results in less noise and a faster response time. 

The noise equivalent power (NEP) of an infrared detector is a measure of the noise generated by the detector and is given by ... 

Figure 3. Broadband spectrum of a conventional 2000 Globar IR source (short dashed line), and the spectrum of the NSLS synchrotron source (solid line) limited by an experimental throughput of 4.4><1 O 4 mm2sr. This is the etendue for a 1 pm by 1 pm sample measured with an infrared microscope. The measured, background limited Noise Equivalent Power (NEP) of a Mercury Cadmium Telluride (MCT) (long dashed line) detector is shown. This detector is operated at liquid nitrogen temperatures.

Material No. on Fig. 3.8 Operating temp. (K) Responsivity (VW- ) Noise equivalent power (WHz Measuring freq. (Hz) Response time (s) Dimensions of element (mm) References/Notes... 

Rather than the noise equivalent power as a figure of merit, the common descriptor for a sampling detector is the noise equivalent electron count, NEE. The output voltage is a measure of the apparent charge transferred to the capacitor, and the three sources of the charge fluctuation may each be categorized as a fluctuation in the electron count, N. The photon noise term is simply (NEE) yot = VN = (rjPlhv)T, from the Poisson statistics of the photoexcitation process. The NEE count associated with the capacitor thermal noise is (NEE)c = = kTC/q, and that of the ampli-... 

Noise equivalent power NEP The incident radiation power generating in the detector an output signal equal to the detector internal noise corresponds to the minimum detectable flux. Measured in W/Hz NEP = - ... 

An FT-IR spectrometer is used optimally when detector noise exceeds all other noise sources and is independent of the signal level. This is the usual case for mid-infrared spectrometry but may not be so for shorter wavelengths. The sensitivity of mid-infrared detectors is commonly expressed in terms of the noise equivalent power (NEP) of the detector, which is the ratio of the root mean square (rms) noise voltage, P , in V Hz to the voltage responsivity, R, of the detector, in V W . It is effectively a measure of the optical power that gives a signal equal to the noise level thus, the smaller the NEP, the more sensitive is the detector. The NEP is proportional to the square of the detector area, Ao, with the constant of proportionality being known as the specific detectivity, D that is. 

Besides the responsivity, the noise properties of a detector are of fundamental importance. The noise is measured in volts (rms) within a specified electrical bandwidth. Since noise in one frequency interval is statistically independent of that in others, the noise power increases linearly with bandwidth and the noise voltage with the square root of the electrical passband. Therefore, the noise characteristic of a detector is expressed in units of V Hz 5. More instructive than the noise voltage per se is the noise normalized to the responsivity. Neither a high responsivity detector with excessive noise nor a low noise element that lacks responsivity is of interest. The noise normalized to the responsivity [V Hz 2 /V W ] is expressed in W Hz 2, and is called the Noise-Equivalent-Power (NEP) per root hertz. In the literature the term NEP (which has units of power, e.g., W) is often applied to the NEP per root hertz (which has units of W 2). This inconsistency is deeply embedded in the literature and we also use the term NEP for both, but we state units where needed to avoid confusion. As the responsivity, the NEP may be a function of wavenumber, the term spectral NEP, NEP or NEP is then appropriate. The NEP [watt] can also be understood as the signal power for a signal-to-noise ratio of unity. The NEP [watt] is generally a small number, the smaller the value the better the detector. The inverse of the NEP [watt] is called the detectivity, D (Jones, 1952),... 

The basic test for any IR detector includes measurement of the output (DC or AC) and noise, as well as the calculations of responsivity and the composite figures of merit noise-equivalent irradiance (NEI), noise-equivalent power (NEP), or D. If an assembly includes only a few elements, and if only a few assemblies are to be tested, it is reasonable to make these measurements with simple voltmeters or wave analyzers, and report the results in a tabular form. Higher volume production requires automated data acquisition equipment and graphical and statistical reporting - much like the testing of FPAs. 

Radiant sensitivity noise equivalent irradiance (NEI) 1 x lO ph/cm /sec (because the applications for this type of instrument measure radiant power rather than temperature, sensitivity is expressed in NEI rather than NETD)... 

Impulse-response and transfer functions can be measured not only by pulse excitation, but also by excitation with monochromatic, continuous waves (CW), and with continuous noise or stochastic excitation. In general, the transformation executed by the system can be described by an expansion of the acquiired response signal in a series of convolutions of the impulse-response functions with different powers of the excitation [Marl, Schl]. Given the excitation and response functions, the impulse-response functions can be retrieved by deconvolution of the signals. For white noise excitation, deconvolution is equivalent to cross-correlation [Leel]. 

Equations (5.24) and (5.25) are equivalent, if one deals with perfect data free of noise. However, if noise is present, as is always the case with measured data, noise contributions computed from Eq. (5.24) are always positive and in the worst case, a factor of s/l larger than the corrected signed noise amplitude computed from Eq. (5.25). The procedure according to Eq. (5.24) and (5.25) is known as the power and Mertz method, respectively. 

---