High Accuracy Modeling Method of UTC-PD Photodiode

Abstract

The high accuracy modeling is a prerequisite to study UTC-PD matching circuit design and power combining. This paper presents the high accuracy modeling method of UTC-PD photodiode. The measured frequency response curve of photoelectric conversion was used as a benchmark, and the model of the equivalent circuit was modified consistently in order to make the frequency response curve of the model fit the measured frequency response curve. This method addressed effectively the modeling issues of UTC-PD photodiode.

SPICE Modeling in Verilog-A for Photo-Response in UTC-Photodiodes Targeting Beyond-5G Circuit Design

“…This implementation, thus, does not meet the criteria for design of high-speed optoelectronic circuits using SPICE. Interestingly, only the phase of the photocurrent is governed by the 𝑒 −𝑗𝜔𝜏 𝑐 /2 term in (9) and without this term, the phase of the photocurrent in (9) reduces to that of the single pole implementation (Fig. 3(c)) even though its magnitude remains the same as that of (9).…”

“…From the compatibility viewpoint for circuit design tools, SPICE models are of crucial importance in photonic device modeling. At present, the existing handful of SPICE-compatible modeling efforts [5][6][7][8][9] in photonics, especially for UTC-photodiodes, are largely based on an equivalent circuit modeling approach. While these equivalent circuit models, consisting of R-C equivalent networks, capture the behaviour of UTC-PDs reasonably well and thus offer computational advantage over numerical models, the physical basis of these models remain limited compared to a physics-based and scalable SPICE compact model in Verilog-A.…”

“…Moreover, this approximation accounts for only one type of carrier in a unipolar region and thus cannot capture the complete physics of carrier transport in devices like p-i-n diodes or UTC-PDs. Considering absorber and collector thicknesses of 100 and 225 nm, respectively, a vth of 2.5×10 5 m/s as well as a vc of 10 5 m/s [10] in (9) for an InGaAs/InP UTC-PDs, the magnitude and phase of the normalized frequency response (Iph/Iph0) is plotted in Figs. 3 (b) and (c) for both the analytical equation and the single pole approximation.…”

“…Interestingly, only the phase of the photocurrent is governed by the 𝑒 −𝑗𝜔𝜏 𝑐 /2 term in (9) and without this term, the phase of the photocurrent in (9) reduces to that of the single pole implementation (Fig. 3(c)) even though its magnitude remains the same as that of (9). From the point of view of design and understanding of microwave photonic systems, capturing both magnitude and phase of the frequency response of photodiodes is crucial yet remains experimentally difficult as reported previously [21][22].…”