Modeling bipolar transistors using multibias S parameter sets

This paper presents a technique for accurate small-signal modeling of microwave bipolar transistors. Model parameters are extracted using measured S parameters in some bias points within the forward active region and then calculated in every point of that region by using a first order interpolation algorithm. The resulting multibias model shows excellent agreement with the measured data over the entire operating region. Moreover, accurate agreement is also achieved with the measured minimum noise figure even though the parameter extraction technique does not make use of noise measurements. Although model parameters are only extracted on measured data at 1 and 5 GHz, scattering parameters are accurately predicted up to 50 GHz, i.e. at a frequency exceeding the transistor FT