Detector Metrics

  • Noise Density is the noise power per unit bandwidth. Noise power is a variance with units of A2 or V2. To relate this to signal from the detector, the square root is taken to get A/√Hz or V/√Hz.

  • Responsivity is the ratio of the detector output signal to the incident radiant power and has units of A/W or V/W. Responsivity is a function of wavelength.

  • Noise Equivalent Power (NEP) is the ratio of noise density to responsivity and has units of W/√Hz. The minimum NEP is sometimes specified, which corresponds to the wavelength of maximum responsivity, λR. The noise equivalent power can then be calculated for a particular wavelength with NEP(λ) = NEPmin * R(λR ) / R(λ). The product of the NEP and the square root of the bandwidth is the signal power required to achieve a signal-to-noise ratio of one.

  • Detectivity (D) is the reciprocal of NEP. The higher the detectivity, the lower the noise.

  • Specific Detectivity (D*), pronounced “dee-star”, is the detectivity multiplied by the square root of the detector area, usually given in Jones units, cm√(Hz)/W. Since D* is independent of size, it is more useful for comparing detector materials than D.

_images/det_dStar.png

Figure 1. Specific detectivity (D*) plots from Judson/Teledyne. Click for higher resolution.

Test Yourself

What is the noise floor of a 0.5 cm square detector with a D* of 1E10 Jones sampled at 1 kHz?

(0.5 cm)*√(1 kHz)/(1E10 Jones) = 1.581 nW