臺灣博碩士論文加值系統

我們提出一個根據雙回授電路法推導出的嶄新理論來解釋電晶體散射參數(即S參數) S22的曲線在史密斯圖(Smith Chart)上所發生的奇特現象。我們在推導的過程中發現，所有電晶體的輸出阻抗在低頻時都呈現出一種電阻與電容的串聯效應，而在高頻時又呈現出一種電阻與電容的並聯效應。而電晶體輸出阻抗這種與生俱來在高、低頻帶出現兩極化的特性造成S22在史密斯圖(Smith Chart)上顯現出曲線由等電阻圓繞向等電導圓的奇特轉折點。我們所建立的模型不但能夠預測S22的行為，還可以準確地計算出另外三個S參數。基於這套理論，我們更進一步衍生出一個可以把小訊號模型參數從量測到的S參數加以淬取出來的簡單方法。
接下來，我們製作了一個完整的電感元件庫，並利用LIBRA軟體針對每一個電感進行模型的建立。這個電感的模型不但使射頻電路設計師在進行模擬時能夠更精確掌握電路的輸出結果，也顯示出電感的Q-Factor是如何受到繞線寬窄、間隔和內圈直徑大小的影響。
在分析了Q-factor的物理意義之後，我們著手進行改善Q值的工程。我們成功地運用微機電式的蝕刻法除去電感下的矽材質基板。在CMOS製程中，晶片式電感的Q值獲得了相當的改善。在2.4GHz下，一個六圈的電感(約11nH)的Q值從3.33被提升至5.19。我們也針對這些底下基板已被蝕刻掉的電感進行模型的建立，被淬取出來的參數顯示出基板阻值大大地上升，而基板電容則大部分都下降了。這代表我們對矽基板的蝕刻確實大大地減少旁生效應所造成的能量損耗。
最後，我們設計了一個低雜訊放大器、一個Double Balanced混波器及一個寬頻放大器來實現通訊系統中的射頻前端裝置。我們在低雜訊放大器中，利用了前面所提到的微機電式電感；並利用電流注入的方式來偏壓我們這個根據Gilbert Cell概念所設計出來的混頻器。而更重要的是，在寬頻放大器的模擬及討論中，我們應用了前面所提出的雙回授電路理論推導出寬頻放大器的電壓增益、輸入及輸出阻抗，以及最大平坦響應之充要條件。
在這本碩士論文當中，我們作了一些貢獻:
 我們提出一個可以簡單並準確計算電晶體S-參數的方法，藉此也提供了一個從低頻數電晶體S-參數淬取其小訊號參數方法。
 我們將微機電式電感這個實驗成功地完成了，並找出來可提升良率的方法。
 我們利用台積電的CMOS製成實現一個Gilbert cell變化電路，藉以完成一個可在射頻前級工作的降頻器。

A novel theory based on dual-feedback circuit methodology is proposed to explain the kink phenomenon of transistor scattering parameter S22. Our results show that the output impedance of all transistors exhibited as a series RC circuit at low frequencies and a parallel RC circuit at high frequencies. It is this inherent ambivalent characteristic of the output impedance that caused the appearance of kink phenomenon of S22 in a Smith chart. Our model not only could predict the behavior of S22 but also could calculate all the S parameters accurately. A simple method for extracting transistor equivalent circuit parameters from measured S parameters was also proposed based on our theory.
A complete inductor library is established and the model for each inductor is extracted by S-parameter curve fitting. The model not only facilitates the RF circuit design but also presents how the size (line width, space, number of turns and the inner diameter) of an inductor influences its Q-factor.
CMOS On-chip inductor with high Q-factor is realized by micromachining. The silicon substrate under inductors is successfully etched and removed. The Q-factor of a 6-turn, 11-nH inductor can be improved from 3.33 to 5.19 at 2.4GHz. The parasitic effects are greatly reduced above 2GHz. The model for each micomachined inductor is also extracted and compared to the normal inductor model. The results show that the substrate resistance is increased and the parasitic capacitance is decreased.
To realize RF front-end of the communication system, a low-noise-amplifier is designed and implemented with micromachined inductors. A double balanced mixer based on the Gilbert Cell topology is implemented with current injection biasing technique. A broadband amplifier based on the dual-feedback theory is designed. The voltage gain, input and output impedance, and the maximum-flat response condition are derived.
In this dissertation, several contributions are made:
 A novel methodology based on dual feedback theory is proposed for people to calculate transistor S-parameters more easily and extract the small signal parameters from S-parameters measured at low frequency.
 A complete experiment on micro-machining on-chip inductors is accomplished.
 A Gilbert cell variant mixer, the micro-mixer, is demonstrated with CMOS technology.

Chapter 1 Introduction 1
1.1 RF in the Bussiness World 1
1.2 RF Device Modeling in CMOS 1
1.3 On-chip Inductors 3
1.4 Building Blocks of Transceivers 4
1.5 Objective of This Work 4
Chapter 2 RF Device Modeling 7
2.1 Deembed Technique 7
2.2 S-parameters of MOSFETs 9
2.3 Extraction of MOSFET's Small Signal Parameters 28
2.4 Spiral Inductor Model 31
Chapter 3 Micromachined On-chip Inductors 37
3.1 Definition of Quality Factor 37
3.2 Design of High-Q Inductor 38
Chapter 4 RF Circuits Design 55
4.1 Low Noise Amplifier 55
4.2 Double Balanced Mixer 63
4.3 BroadBand Amplifier 76
Chapter 5 Conclusion 87
Conclusions 87
Reference
Appendix

[1] David Lovelace, Julio Costa and Natalino Camilleri, “Extracting small-signal
model parameters of silicon MOSFET transistors,” IEEE MTT-S Digest, pp.865~868, 1994.
[2] Chung-Hwan Kim, Cheon Soo Kim, Hyun Kyu Yu and Kee Soo Nam, “Unique
extraction of substrate parameter of common-source MOSFET’s,” IEEE Microwave and Guided Wave Letters, VOL. 9, NO. 3, pp. 108~110, March 1999.
[3] W. Liu, R. Gharpurey, M. C. Chang, U. Erdogan, R. Aggarwal, and J.P. Mattia,
“R.F. MOSFET modeling accounting for distributed substrate and channel resistance with emphasis on the BSIM3v3 SPICE model,” IEEE IEDM, pp.309~312, 1997.
[4] Julio Costa, Dave Lovelace, Dave Ngo, and Natalino Camilleri, “Modeling a new
generation of RF devices: MOSFETs for L-band applications,” IEEE MTT-S Digest, pp.293~296, 1993.
[5] M.C. Ho, F. Brauchler and J.Y. Yang, “Scalable RF silicon MOSFET distributed
lumped element model based on BSIM3v3,” Electronics Letters, Vol. 33, No. 23, pp.1992~1993, November 1997.
[6] C. Patrick Yue, S. Simon Wang, “On-chip spiral inductors with patterned ground
shields for Si-based RF IC’s,” IEEE Journal of Solid State Circuits, Vol. 33, No. 5, pp. 743~752, May 1998.
[7] J. Y. -C. Chang, Asad A. Abidi, and Michael Gaitan, “Large suspended inductors on silicon and their use in a 2-m CMOS RF amplifier,” IEEE Electron Device Letters, Vol. 14, No. 5, pp.246~248, May 1993.
[8] Behzad Razavi, “CMOS RF receiver design for wireless LAN applications,”
Radio and Wireless Conference, pp. 275~280,1999.
[9] David M. Pozar, Microwave Engineering, Second Edition, John Wiley & Sons,
1998, pp. 204~205.
[10] B. Bayrajtariglu, N. Camilleri, S. A. Lambert, “Microwave performance of n-p-n and p-n-p AlGaAs/GaAs heterojunction bipolar transistors,” IEEE Trans. Microwave Theory Tech., vol. 36, no. 12, pp. 1869-1873, 1988.
[11] P. R. Gray and R. G. Meyer, Analysis and Design of Analog Integrated Circuits, John Wiley & Sons, Inc., 1993, pp.579-584.
[12] R. A. Minasian, “Simplified GaAs M.E.S.F.E.T. model to 10 GHz,” Electronics Lett., vol. 13, no. 18, p. 549, 1977.
[13] Y. J. Chan, C. H. Huang, C. C. Weng, and B. K. Liew , IEEE Trans. Microwave Theory Tech., vol. 46, No.5, 1998, pp.611-615
[14] Y. Aoki and Y. Hirano, “High-power GaAs FETs,” in High Power GaAs FET Amplifiers, John L. B. Walker, Ed., Boston: Artech House, 1993, p.81.
[15] M. Fukuta and Y. Hirachi, “Fundamentals of GaAs field effect transistors,” (in Japanese), Electronic Information and Communication Society, 1992, p.81.
[16] C. Patrick Yue,Changsup Ryu, Jack Lau, Thomas H. Lee, and S. Simon Wang, “A physical model for planar spiral inductors on silicon,” IEEE IEDM, pp.155~158, 1996.
[17] Alex Zhenjun Zhu and Chang, University of Illinois at Urbana-Champaign.
[18] K. Sato, M. Shikida, T. Yamashiro, K. Asaumi, Y. Iriye, M. Yamamoto, “Anisotropic etching rates of single-crystal silicon for TMAH water solution as a function of crystallographic orientation,” Micro Electro Mechanical Systems, pp. 556~561, Jan 1998.
[19] Osamu Tabata, “pH-controled TMAH etchants for silicon micronmachining,” Solid-State Sensors and Actuators, Vol. 1, pp. 83~86, June 1995.
[20] K. Lian, B. Stark, A.M. Gundlach, A.J. Walton, “Aluminium passivation for
TMAH based anisotropic etching for MEMS applications,” Electronics Letters, Vol. 35, pp. 1266~1267, July 1999.
[21] Thomas H. Lee, The Design of CMOS Radio-Frequency Integrated Circuits, Cambridge.
[22] Gonzales, Design of Microwaves Circuits, chap4.
[23] Barrie Gilbert, "The MICROMIXER: A highly linear variant of the Gilbert mixer using a bisymmetric Class-AB input stage," IEEE Journal of Solid State Circuit, Vol. 32, pp. 1412~1423, Sept. 1997.
[24] Jeff Durec and Eric Main, “A liner class AB single-ended to differential transconverter suitable for RF circuits,” IEEE MTT-S Digest, pp. 1071~1074, 1996.
[25] Leonard A. MacEachern and Tajinder Manku, “A charge-injection method for Gilbert cell biasing,” Electrical and Computer Engineering, 1998. IEEE Canadian Conference, pp. 365 ~ 368, May 1998.