Analysis of an optoelectronic mixer based on metal‐semiconductor‐metal photodetector

Abstract

In this article we present the analysis of an optoelectronic mixer (OEM) based on a metal‐semiconductor‐metal photodiode (MSM‐PD) in microwave fiber‐optic links. The nonlinearities inherent in the MSM‐PD OEM are examined by modeling its dc responsivity as a third‐order polynomial function of applied bias voltage. The OEM has been analyzed in terms of its conversion loss, carrier‐to‐noise ratio, and noise figure. © 1995 John Wiley & Sons, Inc.

An improved analysis of an optoelectronic mixer based on a metal-semiconductor-metal photodetector

“…(8)From (5)-(8) one can see that in the absence of the LO voltage; that is, VLo = 0, (5) turns out to be the dc responsivity. This result agrees well with the results reported in[5]. However, in contrast with the observations of[5], from (6)-(8) one cannot see that the coefficients K , become less dependent on V,, as 1 increases.…”

“…where I = odd integer, (7) and 2(2n + l)a where Pa is the optical power incident on the detector, r is the optical modulation index, w, is the input signal modulation frequency and R ( t ) is the responsivity expressed by (5). then the parameters L,, CNR,,,, CNR,,,.…”

“…Liu, and MacDonald [5] derived the model of (1) for the dc responsivity of the MSM-PD OEM based on the experimental measurement of the responsivity as a function of the dc ANALYSIS Figure l(a) shows the measured dc responsivity of an MSM-PD as a function of the bias voltage Vdc [5]. Here we propose representing this characteristic with the use of the Fourierseries model of equation (2):…”

“…Thus ( I 1 reduced to a simple polynomial equation with N = 3. With the use of the resulting third-order polynomial, expressions were obtained for L c , CNR,,,, CNR?,, and F,, [5].…”

“…However, because of the hyperbolic tangent function, tanh( aVdc), (1) in its present form cannot yield closed-form expressions for the conversion loss ( L J , the carrier-to-noise ratio of the upconverted signal (CNR,,,), the carrier-to-noise ratio of the input signal (CNRi,,), and the noise figure (4,). Therefore, Prasad et al [5] focused their interest on the case where the argument of the hyperbolic tangent function is large enough. Thus ( I 1 reduced to a simple polynomial equation with N = 3.…”