(Invited) Physics and Compact Modeling of SiGe HBT Linearity Using Mextram

Abstract

This paper presents fundamentals of SiGe HBT RF linearity and its compact modeling using Mextram 504.12 and the latest Mextram, 505.00, for both common-emitter and common-base configurations.Collector-base junction depletion capacitance model is shown to be significant for accurate modeling of peak IP3 behavior, for which two new options are introduced in Mextram 505.00.Current dependence of avalanche factor is shown to be important for accurate modeling of IP3 peak. This is increasingly important in emerging RF applications requiring SiGe HBTs to operate at increasingly high VCB with negative IB .

On the Mechanisms of Linearity Peak Modeling in Common-Emitter SiGe HBTs Using Mextram 505

“…However, as recently shown in [14], V B 2 C 2 is indirectly related to V * B 2 C 2 via I N , and includes some effects of velocity saturation. V B 2 C 2 was shown in [14] to be a better choice for V junc than another BC voltage in Mextram, V * B 2 C 2 , the effective BC voltage for calculation of main current I N , as the value is physically much more meaningful at low current than V * B 2 C 2 , which is mainly a result of mathematical smoothing of the epi-layer injection width x i before onset of quasi-saturation. Such effects, however, enter V * B 2 C 2 , and hence indirectly influence…”

“…Model parameters are extracted by fitting measured DC I-V, S-parameters, and IP3 characteristics. As discussed in [14], Mextram 504 has problem modeling measured IP3 peak behavior. New options are developed in Mextram 505 to improve IP3 peak modeling.…”

“…Historically, in Mextram 504, V B 2 C 2 was not considered proper for use as BC junction depletion charge control voltage, as the Kull model is used to formulate I C 1 C 2 , the epilayer current, which does not account for velocity saturation [15]. However, as recently shown in [14], V B 2 C 2 is indirectly related to V * B 2 C 2 via I N , and includes some effects of velocity saturation. V B 2 C 2 was shown in [14] to be a better choice for V junc than another BC voltage in Mextram, V * B 2 C 2 , the effective BC voltage for calculation of main current I N , as the value is physically much more meaningful at low current than V * B 2 C 2 , which is mainly a result of mathematical smoothing of the epi-layer injection width x i before onset of quasi-saturation.…”

X-Parameters Based Characterization and Compact Modeling of SiGe HBT Linearity

“…For relatively small signal RF circuits such as LNAs, linearity is often characterized by measuring 3rd order intercept points, IP3, using a spectrum analyzer, or better, a Nonlinear Vector Network Analyzer (NVNA), which also provides phase information. For highly linear common-base SiGe HBTs [1], however, we find that the distortions generated within the RF source, which itself is also a transistor circuit, can be amplified by the HBT under test, and then interfere with the HBT's distortions in a vector manner. The measured IP3 is inaccurate when such interference is significant, typically at biases of high IP3, which is of great interest to circuit design.…”