Inp DHBT Device Modeling And Circuit Design

Due to the ever increasing bandwidth requirements of wireless communications systems, the large amount of spectrum available at millimeter-wave frequencies is making these frequency bands increasingly more significant. This spurs the need for the development of the millimeter wave circuit and system building blocks. With both high efficiency and high lineaiity capability, InP HBT technology has an irreplaceable role in the millimeter wave transceiver system. Though not currently suitable for high-density integration digital logic, these devices are ideal for advanced millimeter wave circuit design and millimeter wave system because of their ability to efficiently generate and amplify high-frequency signals.Millimeter wave measurements pose serious challenges during the HBT modeling cycle. Both the DC and RF measurement techniques used for the devices measurements.In this paper, we have conduted the DC and small signal characteristics test of InP HBT from DC to 220 GHz. The DC characteristics and s-parameters of InP HBT are measured using an Agilent 4156 and N5247A network analyzer (up to 70 GHz), millimeter wave controller N5262A, VNA (vector network analyzer) frequency extenders from Farran Technology of series FEV-10-TR (75-110 GHz), FEV-05-TR (140-220 GHz) with a probe station from Cascade Microtech (Summit 12000m), respectly. And complete the InP HBT small signal modeling in the frequency range of DC-220 GHz. On the basis of the completion of the small signal model parameter extraction, Completion of DC-67 GHz large signal modeling based on the characteristics of Agilent FIBT model followed by implement the large signal model to ADS for millimeter circuit design.The main research results are listed as following".1) The commonly used calibration methods SOLT, LRRM, LRM and TRL are discussed. Due to the shortcomings of the SOLT,LRM, LRRM has obvious advantages in broadband test, and is supported by Cascade Microtech’s WinCal software. The match standard of LRRM is assumed as a resistance and inductance in series, this is advantage over LRM calibration methods.2) The advanced de-embedding methods are presented in this work. The differences among them are listed and the results of the most commonly used methods are given in the last.3) The errors of VNA are corrected by using LRRM calibration. With S-parameters measurement of Open dummy structure, the boardband modeling of Open structure is achieved in 0.2-220 GHz. For accurate small signal modeling, we propose the new de-embedding method. The equivalent circuit topology of Open structure modeling 0.2-220 GHz is used to achieve 0.2-220 GHz wideband InP HBT small signal model. Excellent agreement between measured and modeled are obtained.4) Once the small signal model of HBT are obtained, the external resistances are achived and directly used for large signal model parameter extraction. Finally, the HBT large signal model is implemented into the ADS software for millimeter circuit design. Factors that have a major impact on the phase noise are analyzed, and some measures are taking to improve the phase noise of VCO.