Font Size: a A A

LTCC Work With Baron Design

Posted on:2010-06-18Degree:MasterType:Thesis
Country:ChinaCandidate:Q LiFull Text:PDF
GTID:2208360275482757Subject:Electromagnetic field and microwave technology
Abstract/Summary:PDF Full Text Request
The LTCC passive microwave components are now being widely used in communications radars and other areas because of small size, high reliability and low cost. The structure of LTCC passive microwave components are muti-layer, and the electromagnetic coupling of the layers is complex. Circuit-theory-based simulation and CAD tools using empirical device models are fast: analytical solutions or available exact derivatives may promote optimization convergence but its accuracy can not be guaranteed. EM simulators has long been used for design verification and also be used for optimize process recently. However, the higher accuracy of the simulation, the more expensive direct optimization cost is expected to be. Although the solutions of EM simulators are accurate, it's difficult to be directly applied to complex EM problems because of the large CPU intensive.Neural space mapping (NSM) is a novel optimization concept, allowing expensive EM optimization to be performed efficiently with the help of fast and approximate models called as coarse models. Through the construction of a mapping between the parameter spaces of the coarse and fine models, a suitable surrogate is obtained. This surrogate is faster than the fine model and at least as accurate as the underlying coarse model. The NSM approach continuously updates the surrogate to approximate the corresponding fine model better.In this thesis, we are absorbed in the design of LTCC power divider and LTCC balun; we verify the feasibility and accuracy of nearal space mapping through the design examples.Firstly, we introduce two rameworks of LTCC capacitances including MIM (Mental-Insulator-Mental) and VIC (Vertically-Interdigitated-Capacitor) and make comparison between them. Based on VIC framework, a LTCC multilayer capacitance has been proposed, and the optimization parameters have been optimized with neural space mapping (NSM). The characteristic of four kinds LTCC inductances have been described.Based on LTCC capacitance and inductance, a novel lumped LTCC power divider has been proposed. Optimization parameters of the LTCC power divider have been obtained in the iterative optimization process between fine model and coarse model, which has been established from multilayer element kit in the ADS software. We improve the parameters again to meet the requirement of the production process, and tape-out the LTCC power divider by use of the production lines of RIEST of UESTC. The test results of the LTCC power divider are basically consistent with the design target.Finally, we present a miniaturized LTCC Balun which work in the 1.7~2.2GHz. We reduce the size of Marchand Balun by the LTCC multilayer technology. There are 11 layers in the LTCC Balun, and the test results of the LTCC Balun which is taped-out basically hit a target.
Keywords/Search Tags:Space mapping, LTCC, neural network, power divider, balun
PDF Full Text Request
Related items